running head: hierarchy in the eye of the beholder...
TRANSCRIPT
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
1
Hierarchy in the Eye of the Beholder: (Anti-) Egalitarianism Shapes Perceived
Levels of Social Inequality
Nour S. Kteilya, Jennifer Sheehy-Skeffingtonb, Arnold K. Hoc
a Department of Management and Organizations, Northwestern University
b Department of Psychology, London School of Economics and Political Science
c Department of Psychology, University of Michigan
Manuscript in press at Journal of Personality and Social Psychology
Corresponding Author:
Nour Kteily ([email protected])
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
2
Abstract
Debate surrounding the issue of inequality and hierarchy between social groups has
become increasingly prominent in recent years. At the same time, individuals disagree
about the extent to which inequality between advantaged and disadvantaged groups
exists. Whereas prior work has examined the ways in which individuals legitimize (or
delegitimize) inequality as a function of their motivations, we consider whether
individuals’ orientation towards group-based hierarchy motivates the extent to which they
perceive inequality between social groups in the first place. Across 8 studies in both real-
world (race, gender, and class) and artificial contexts, and involving members of both
advantaged and disadvantaged groups, we show that the more individuals endorse
hierarchy between groups, the less they perceive inequality between groups at the top and
groups at the bottom. Perceiving less inequality is associated with rejecting egalitarian
social policies aimed at reducing it. We show that these differences in hierarchy
perception as a function of individuals’ motivational orientation hold even when
inequality is depicted abstractly using images, and even when individuals are financially
incentivized to accurately report their true perceptions. Using a novel methodology to
assess accurate memory of hierarchy, we find that differences may be driven by both
anti-egalitarians underestimating inequality, and egalitarians overestimating it. In sum,
our results identify a novel perceptual bias rooted in individuals’ chronic motivations
towards hierarchy-maintenance, with the potential to influence their policy attitudes.
Keywords: Social Dominance Orientation; Hierarchy; Inequality; Motivated Cognition;
Egalitarianism
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
3
The topic of inequality between social groups has become one of the most
prominent socio-political issues of our time. In the wake of events such as the global
recession in 2008, the election of Barack Obama as U.S. President, the recent spate of
fatal police shootings of African Americans, Nobel Laureate Tim Hunt’s comments about
women in science, and prominent reports about contemporary levels of gender
discrimination, public debate surrounding the extent of inequality between groups, and
whether it constitutes a social problem needing to be politically addressed, has increased
noticeably (Baker & Fausset, 2015; Blow, 2013; Guardian, 2015; Grant & Sandberg,
2014; New York Times, 2011, 2013; Williams & Ceci, 2015).
That discussion about social inequality has risen, however, does not imply that
individuals agree about its importance or even its prevalence. As with many of the issues
currently occupying the American political agenda (e.g., climate change and immigration
reform), there is a great deal of polarization about whether there is a problem to be
addressed: that is, about whether groups across the social hierarchy in the U.S. already
enjoy equal opportunities, prospects for social mobility, and access to resources
(Chambers, Swan, & Heesacker, 2015; Cohn, 2014; Kraus & Tan, 2015; Porter, 2014).
Some, such as members of the “Occupy” grassroots social movement that arose
following the 2008 recession, have vocally denounced what they perceive to be gross and
structurally-sustained inequalities in American society. Others, such as supporters of the
Tea Party, have labeled such claims as unfounded, focusing instead on findings
suggesting that social mobility is as strong as it ever has been in the U.S. (Chetty,
Hendren, Kline, Saez, & Turner, 2014), while ethnic and gender diversity in high-paying
jobs has vastly improved (Barak, 2013).
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
4
At stake in these debates is the prospect of egalitarian social change. Because the
existence of extensive intergroup hierarchy is considered a problem in a society
professing egalitarian social norms (e.g., the United States; Katz & Hass, 1988), drawing
attention to wide disparities between groups at the top and groups at the bottom raises the
specter of the need for redistributive policies (such as progressive taxation or affirmative
action) that seek to bring about greater equality between social groups (e.g., Brodish,
Brazy, & Devine, 2008; Lowery, Knowles, & Unzueta, 2007). Because of its potential
implications for support of egalitarian social policy, the degree of inequality in society
may become a point of political contention between those who long for a world of
complete equality (like many Occupy supporters) and those who favor the maintenance
of a hierarchical social order (like many followers of the Tea Party).
In the current work, we advance the possibility that individuals’ motivations to
maintain or attenuate hierarchy influence the level of social inequality they perceive.
Drawing from research on motivated cognition, we reason that those who seek to
attenuate hierarchy in society come to see the gross inequities that justify the belief that
egalitarian social change is needed, whereas those who seek to maintain hierarchy come
to perceive society as relatively equal, and thus that egalitarian policies are unnecessary.
Motivations for Hierarchy-maintenance vs. Hierarchy-attenuation
One variable that captures variation in individuals’ support for social hierarchy
between groups is social dominance orientation (i.e., SDO; Ho et al., 2015). Whereas
individuals high in SDO seek to maintain hierarchy between groups in society, those low
in SDO favor intergroup equality. One previously explored mechanism by which high
(vs. low) SDO individuals resist (vs. push for) egalitarian social change is through
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
5
endorsing ideologies that justify (vs. problematize) existing levels of inequality between
groups. High SDO individuals favor ideologies such as the Protestant work ethic,
classism, racism, sexism, specieism, or karma that can help justify existing inequality
(Bobocel, Son Hing, Davey, Stanley, & Zanna, 1998; Cotterill, Sidanius, Bhardwaj &
Kumar, 2014; Dhont, Hodson, Costello, & MacInnis, 2014; Ho et al., 2012, 2015;
Knowles & Lowery, 2012; Kteily, Ho, and Sidanius, 2012; Sibley, Wilson, & Duckitt,
2007; Sidanius, Cotterill, Sheehy-Skeffington, Kteily, & Carvacho, 2015). These
ideologies, termed hierarchy-enhancing legitimizing-myths, help provide a legitimate
explanation for why some groups are doing better than others in society: If groups’
outcomes can be said to be unequal because of legitimate differences in effort or ambition
(as opposed to unequal access to opportunities), group-based disparities are less likely to
stoke calls for redistributive social policy (Sidanius & Pratto, 1999).
Although less research has investigated these phenomena from the perspective of
those who seek to bring about equality, low SDO individuals, by the same token, endorse
hierarchy-attenuating legitimizing myths that serve to promote the need for egalitarian
policy. Specifically, they favor ideologies— such as the belief that healthcare, education,
or living wages are universal human rights— that problematize existing social inequality
(Pratto, Sidanius, & Levin, 2006). To the extent that such ideologies prosper, any existing
group-based inequality in these domains, no matter how small, tends to become a social
concern requiring intervention.
Thus, even when they perceive identical levels of inequality between groups in
society, those motivated to maintain vs. reduce hierarchy may explain it differently, in
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
6
ways that either downplay or prioritize it as a social issue worthy of attention and
correction.
SDO and Motivated Perception of Hierarchy
Another complementary possibility, heretofore largely unexamined, is that
individuals may actually perceive the degree of inequality differently as a function of
their motivations, with those motivated to achieve equality perceiving a lot of social
hierarchy, and those motivated to maintain inequality perceiving less. We reason that this
pattern of biased inequality-perception may result from individuals’ motivated responses
to social pressures. In societies with egalitarian norms, such as the U.S., recognition of
extensive levels of hierarchy generates calls for greater attention and explanation (Bobo
& Kluegel, 1993; Murrell, Dietz-Uhlrer, Dovidio, Gaertner, & Drout, 1994; Lowery et
al., 2007). Because those who want to bring about equality are aware that social support
for egalitarian policies increases the more inequality there is, they may be driven to
perceive more hierarchy in society, justifying their desire for action. For the converse
reason, those motivated to maintain hierarchy may be driven to perceive less of it,
justifying their desire to avoid egalitarian intervention. Here, we draw on the growing
body of literature on the motivated nature of cognition, information processing, and
perception to examine the novel possibility that equality-related motivations can
influence the perception of hierarchy between social groups.
What functions might motivated perceptions of hierarchy serve? Insights from
research on motivated cognition suggest that motivated perceptions of hierarchy could
have both psychological and practical benefits: First, perceiving the level of hierarchy
that fits (vs. conflicts) with one’s views on the desirability of achieving social equality
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
7
could reduce any psychological dissonance that might otherwise be experienced (see
Balcetis & Dunning, 2007). Second, perceiving a level of hierarchy consistent with one’s
motivations may redouble one’s commitment to their goal of implementing (vs. avoiding)
egalitarian social policies, as well as bolstering their ability to convincingly persuade
others of their perspective (Von Hippel & Trivers, 2011). For example, when individuals
who favor equality come to actually perceive greater power gaps between groups, they
should see an even greater need to implement specific egalitarian social policies (which
aim to minimize group-based differentials); this should then manifest both in their own
heightened support for these policies, as well as greater dedication and effectiveness in
convincing others of their views.
Below, we review evidence outlining how social dominance motives might come
to influence perceptions of the extent of power differences between groups.
Mechanisms linking SDO to Biased Perception of Inequality
Thus far, we have reasoned that beyond influencing the types of ideologies
individuals adopt in order to justify or problematize the existing level of hierarchy, SDO
motives may also influence the degree of hierarchy individuals perceive. Specifically, we
argued that low SDO individuals might be motivated to maximize power differences
between groups, whereas high SDO individuals might be motivated to minimize them.
We now consider how precisely individuals might come to perceive hierarchy differently
as a function of their social dominance motives.
The influence of motivation on the maximizing versus minimizing of inequality
between social groups could occur at two stages: initial perception of inequality, or later
reporting of it. From the perspective of a motivated reporting account, individuals high
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
8
and low in SDO might perceive the same gaps in power between advantaged and
disadvantaged groups, but, because of their motivations towards vs. against hierarchy,
strategically and consciously downplay or exaggerate them (respectively) in discussions
with others.
From the perspective of a motivated perception account, on the other hand,
individuals’ motivations might influence their actual underlying perception of group-
based differences in power and resources. According to this view, rather than reflecting
conscious selective reporting, SDO motives drive individuals to perceive fundamentally
differing social realities, potentially outside their conscious awareness. If individuals
truly see the world differently as a function of their motivations towards equality, this
would suggest that they have hierarchy-enhancing vs. attenuating perceptual biases. Such
biases could serve a function similar to that of hierarchy-enhancing vs. attenuating
legitimizing ideologies— bolstering individuals’ beliefs about whether egalitarian social
policies are needed— but in a different way (i.e., by changing their perception of the
world rather than their explanations for it). The idea that SDO motives can influence
social perceptions, with implications for maintaining hierarchy, has received some prior
support. For example, Whites higher (vs. lower) in SDO perceive low status ambiguous
targets (e.g., unemployed Black-White biracials; Dzhokhar Tsarnaev) as looking less
White, thus excluding them from the high status group and helping to maintain intergroup
differentiation (Kteily, Cotterill, Sidanius, Sheehy-Skeffington, & Bergh, 2014; see also
Ho, Sidanius, Cuddy, & Banaji, 2013).
By what mechanisms might individuals’ equality-motives lead them to have truly
different perceptions of hierarchy in society? There are at least two related possibilities
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
9
that we consider in turn: First, individuals’ motivations might lead them to self-select into
different environments in which they end up being inherently more vs. less exposed to
information about power differences. Second, even where individuals encounter identical
underlying information about power differences, their motivations may lead them to
process it differently: Thus, consistent with motivated cognition, individuals may filter
information in a biased manner, for example judging the validity or importance of
information about power differences differently, or selectively focusing on the aspects
that confirm their worldviews.
According to a differential exposure account, individuals high and low in SDO
may end up occupying different social worlds. Because of their aversion to low status
targets (e.g., Duckitt & Sibley, 2010; Kteily et al., 2014; Thomsen, Green, & Sidanius,
2008) and their preference for hierarchy-enhancing jobs (Sidanius, Pratto, Sinclair, &
Van Laar, 1996), individuals higher in SDO may be more likely than those lower on SDO
to avoid certain types of social contexts (e.g., poor neighborhoods) or occupations (e.g.,
social worker; public defender), which happen to provide more information highlighting
discrepancies between those at the bottom and those at the top (see Dawtry, Sutton, &
Sibley, in press; Garrett, 2009). Conversely, and for the opposite reasons, low SDO
individuals might be drawn to the types of contexts in which they end up regularly
encountering information about large group-based disparities. If individuals high (vs.
low) on SDO select into environments where they get exposed to different underlying
information about the world, that could account for why they end up perceiving more (vs.
less) hierarchy between groups. Notably, from this perspective, SDO could be negatively
associated with inequality perceptions even if it does not bias how individuals process
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
10
any hierarchy-relevant information that they are exposed to (e.g., a newspaper article
about rising income inequality).1
Research on motivated cognition and perception, however, suggests that there are
good reasons to expect that individuals’ equality motives will indeed bias their processing
of hierarchy-relevant information. Specifically, individuals’ motives may lead them to
filter information differently, in ways that sway them to come to the conclusions they
prefer. As we detail below, biased information filtering could manifest in a number of
ways.
For example, individuals may pay more attention to information that is consistent
with their motives (and less to information that is inconsistent). Research drawing on data
from millions of online interactions on social media platforms (e.g., Twitter; Facebook)
shows behavioral evidence for this idea: Specifically, individuals are more likely to click
on (and share) online content the more it aligns with their political beliefs (Bakshy,
Messing, & Adamic, 2015; Barberá, Jost, Nagler, Tucker, & Bonneau, 2015), suggesting
a greater tendency to engage with (and endorse) worldview-consistent perspectives. Low
SDO individuals, who consider inequality problematic, might be especially attuned to
any evidence of even slight disadvantage to low power groups. On the other hand, high
SDO individuals, who consider hierarchy desirable, might expend less effort looking for
evidence of low power group disadvantage and more time focusing on any threatening
advances they might be making (Eibach & Keegan, 2006; see Balcetis & Dunning, 2013
for a discussion of the role of attention in motivated perception). 1 Our use of the term ‘differential exposure’ should be differentiated from the term ‘selective exposure’. Whereas the latter term is typically used to describe motivated processing of information (e.g., Frey, 1986; Hart et al., 2009)— the mechanism we consider next— the former reflects the possibility that individuals might simply happen to encounter different underlying information about hierarchy, resulting in different perceptions of hierarchy which do not necessitate biased processing of information.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
11
Beyond guiding which information is attended to, individuals’ motives may
influence how that information is interpreted. For example, individuals may differentially
judge the validity of information about inequality that they come across. Individuals are
known to apply less rigorous standards when evaluating information in line (vs.
inconsistent) with their worldview (Ditto & Lopez, 1992), judging evidence consistent
with their beliefs as more legitimate than evidence that challenges them (Ditto, Pizarro, &
Tannenbaum, 2009; Lord, Ross, & Lepper, 1979; Munro & Ditto, 1997; see also Taber
and Lodge, 2006). Of direct relevance to the present work, Crawford, Jussim, Cain and
Cohen (2013) showed that SDO influenced individuals’ evaluations of the validity of
newspaper article espousing pro vs. anti-affirmative action articles, with lower levels of
SDO associated with perceiving more author bias for the anti-affirmative action article
but less author bias for the pro-affirmative action article (see also Iyengar & Hahn, 2009).
Beyond judging the validity of information differently, individuals’ motives may
lead them to consider it from differing vantage points, impacting the conclusions that
they draw. For example, individuals high in SDO experience a closing gap between
dominants and subordinates as threatening and aversive (Guimond, De Oliviera,
Kamiesjki, & Sidanius, 2010; Thomsen et al., 2008) because it renders the hierarchy that
these individuals favor potentially precarious. Interpreting information about drops in
advantaged groups’ power by fixating on the threat it implies to hierarchy may have the
effect of cognitively magnifying the losses. In contrast, a low SDO individual observing
the same drop in advantaged groups’ power might instead encode it as insignificant
because they contrast it to all the other advantages enjoyed by those at the top. Consistent
with the idea that individuals’ motives lead them to apply different interpretive frames to
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
12
hierarchy-relevant information, Eibach and Ehrlinger (2006) showed that White
Americans (presumably motivated to maintain inequality) anchor on how far race-
relations have come whereas Black Americans (presumably motivated to achieve greater
equality) anchor on how far they still have to go.
If low and high SDO individuals differentially attend to and interpret information
that, respectively, highlighted vs. diminished power differentials between groups, they
could come to diverge in their perception of the extent of those power differentials.
Evidence for Motivated Cognition
Recent research provides compelling illustrations of the ways in which motivated
cognition can bias individuals’ perceptions. In one study, Kahan, Hoffman, Braman,
Evans, and Rachlinski (2012) showed individuals the same video of a protest in which
the police used crowd-control tactics. These authors showed that individuals with
egalitarian vs. inegalitarian values diverged in their ratings of how aggressively the police
treated protesters depending on whether they were that told the demonstrators were
protesting in opposition to abortion or in favor of gays’ entry to the military. Balcetis and
Dunning (2006) showed that individuals were more likely to perceive the same
ambiguous shape as a “B” when “13” was associated with an undesired outcome, but as
“13” when the outcome assignments were switched, providing evidence that individuals’
motivations can exert top-down influences even on individuals’ visual perception.
Documenting effects of motivated cognition on memory, recent work showed that
individuals motivated to deny climate change were more likely to recall information in an
article or video describing human contributions to climate change in system-exonerating
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
13
ways (Hennes et al., under review; see also Story, 1988; Sanitioso, Kunda, & Fong,
1990).
Of particular relevance to the current research, two papers examined motivated
evaluations of income inequality (Chambers, Swan, & Heesacker, 2014) and economic
mobility in the U.S. (Chambers et al., 2015) among liberals and conservatives. These
authors compared evaluations against objective metrics of these phenomena, allowing
them to determine whether one or both groups made inaccurate evaluations consistent
with their motivations. Chambers et al. (2014) found that, on average, Americans
overestimated the rise of income inequality over time, and underestimated average
incomes. Moreover, they found that political liberals, for whom rising inequality is
worldview-consistent, overestimated it to a greater extent than conservatives (but see
Erikson & Simpson, 2012; Norton & Ariely, 2011). Similarly, Chambers et al. (2015)
found that, on average, Americans underestimated economic mobility, an effect that was
moderated by ideology (but see Kraus & Tan, 2015). Supporting the bidirectional
perspective that both liberals and conservatives exhibit motivated biases, they observed
that liberals were generally more likely to underestimate economic mobility (particularly
upward mobility), but that conservatives exhibited a tendency to underestimate
downward mobility.
In sum, we argue that in a society with egalitarian norms, where large disparities
between groups call for attention and potential intervention, one’s motivation to maintain
vs. attenuate intergroup hierarchy (i.e. one’s level of SDO) may influence one’s
description of the extent of inequality or social hierarchy. This is more than strategic
reporting, in which low SDO and high SDO individuals perceive the same level of
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
14
hierarchy, but consciously overblow or minimize it in their descriptions of it, in order to
respectively promote or undermine egalitarian social policy. Rather, we predict that
individuals high vs. low in SDO truly come to perceive different levels of hierarchy. This
differential perception of inequality could, in turn, reflect differential exposure to
information about inequality and hierarchy, such as through self-selection into different
social environments that vary in how much information about inequality they feature.
Alternatively, in line with research on motivated social cognition, individuals may come
to process the same information about hierarchy differently, actually perceiving
inequality in ways that align with their desire to see a world in which hierarchy is or is
not a social problem requiring intervention.
Hypotheses
Based on the reasoning above, we made several predictions. Our central
hypothesis was that SDO would be associated with perceptions of inequality between
groups, with high SDO individuals perceiving less hierarchy and low SDO individuals
perceiving more hierarchy. Because SDO is a variable theorized to index support for
hierarchy between groups in general (Kteily et al., 2012; Sibley & Liu, 2010), we
expected that individuals’ SDO would drive their perception of inequality across
domains: thus, higher levels of SDO should be associated with perceiving smaller gaps in
power between not only the rich and poor, but also between Whites and Blacks, and men
and women. Moreover, given that SDO is associated with support for hierarchy among
both advantaged and disadvantaged group members (Ho et al., 2015), we predicted that
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
15
any association of SDO with perceptions of hierarchy would hold among both high and
low power group members.2
We also examined the mechanism underlying any link between SDO and
hierarchy perceptions. Given findings emerging from research on motivated cognition,
we predicted that the relationship might extend beyond conscious strategic under- or
over-reporting of power differences, or differential access to information about hierarchy.
Specifically, we reasoned that perceptions of power differences might reflect an
unconscious, motivated perceptual bias: Thus, even when presenting individuals with
identical information about hierarchy and incentives to report their true perceptions, we
predicted that SDO would be negatively associated with the extent of perceived power
differentials between groups.
Recent research has given conflicting accounts of whether distorted cognition is
equally a feature of the political right and the left: Whereas some research suggests that
those on the right may be more dogmatic and biased in their information processing (e.g.,
Jost, Glaser, Kruglanski, & Sulloway, 2003; Barberá et al., 2015), other research suggests
that biased cognition equally affects both those on the left and right (e.g., Brandt, Reyna,
Chambers, Crawford, & Wetherell, 2014; Kahan et al., 2012). Although this research has
generally focused on liberalism vs. conservatism broadly construed (rather than on SDO
in particular), because of its inconclusiveness with respect to the issue of distorted
2 Notably, neither the prediction that SDO will lead to perceptions of hierarchy across domains nor that it will do so among both members of advantaged and disadvantaged groups follows from theorizing focusing exclusively on group-membership based motives (e.g,. Tajfel & Turner, 1979; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). According to that view, perceptions of hierarchy should be restricted to whatever context is relevant to the salient ingroup, and individuals should be driven primarily by the motive to advance ingroup interests.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
16
cognition, we made no strong predictions regarding whether the relationship between
SDO and perception of hierarchy would be driven more by those high or low in SDO.
Extending prior work, we also examined the relationship between perception of
hierarchy and policy support. Because egalitarian social policies are specifically intended
to equalize conditions between groups, those who perceive greater inequality should see a
more pressing need to adopt these policies than those who perceive less. Thus, we posit
an indirect path whereby individuals’ general motivations to maintain vs. attenuate
hierarchy (i.e., their SDO) will influence their perception of the level of hierarchy in
specific contexts, which contributes to support for vs. opposition to egalitarian social
policies in those contexts.
In examining these predictions, we controlled for social conservatism. Theorizing
on political conservatism (Jost et al., 2003; Jost, Federico, & Napier, 2009), suggests that
it reflects two core motivations: traditionalism, reflecting a desire to resist change, show
deference to authority and maintain the social order, and acceptance of inequality,
reflecting anti-egalitarianism. These two components are correlated but theoretically
distinguishable; whereas traditionalism is primarily reflected in social conservatism,
acceptance of inequality is directly tapped by SDO (Duckitt, 2001). Because our
predictions follow specifically from motivations to maintain (vs. attenuate) inequality in
society, we examined the effects of SDO controlling for social conservatism in our
analyses.
Overview of Studies
We explored the relationship between anti-egalitarianism and perceptions of
hierarchy in eight studies (and four supplemental studies), involving a range of real-world
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
17
social contexts, fictitious vignettes, and abstract visual representations of hierarchy. In
our first two studies, we examined, cross-sectionally (Study 1a) and longitudinally (Study
1b), the extent to which SDO is linked to the rejection of egalitarian social policy through
diminished perception of power differences among real-world groups, focusing on
perceived disparities between racial, gender, and class groups (see also Supplemental
Study 1). In Study 2, we controlled for the possible effects of differential environmental
exposure to hierarchy-relevant information by examining ratings of a fictional
asymmetric intergroup conflict described using identical text (see also Supplemental
Study 2). In Studies 3a and 3b, we examined the relationship between SDO and
perceptions of hierarchy in a different modality, by using abstract visual depictions of
unequal distributions of power and resources (see also Supplemental Study 3). In Study
3c, we examined whether the relationship between SDO and perceptions of hierarchy
extended to other measures of egalitarianism. In Study 4, we experimentally examined
the extent to which the relationship between SDO and perceived power differences is
sensitive to monetary incentives for accuracy, providing clarity about whether results are
due to motivated reporting or motivated perception (see also Supplemental Study 4).
Finally, in Study 5, we examined whether SDO predicted biased memory, assessing both
whether individuals high in SDO were likely to make errors consistent with under-
perceiving hierarchy, and whether individuals low in SDO were likely to make errors
consistent with over-perceiving it.
Study 1a
In Study 1a, we examined the relationship between SDO and perception of power
differences across three real-world domains: race, gender, and social class. Specifically,
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
18
we asked participants to report their rating of the power enjoyed by groups across these
three intergroup contexts. In order to examine our mediational hypothesis (i.e., that SDO
would predict perceptions of inequality between groups in society, thus influencing social
policy support), we also assessed endorsement of a broad set of egalitarian social policies
spanning a range of domains. Among this set were policies specific to the intergroup
contexts on which we obtained inequality ratings. However, because SDO is a variable
indexing generalized attitudes towards hierarchy (Kteily et al., 2012; Sibley & Liu,
2010), we also examined the generalizability of our hypothesized mediation by including
more general hierarchy-enhancing and attenuating policies (i.e., assessing whether SDO’s
hypothesized link with hierarchy perception relates to endorsement of egalitarian social
policies across a range of contexts). Finally, to assess our prediction that these effects
would hold across the spectrum of membership in high vs. low power groups, we
examined whether participants’ race, gender, or social class membership moderated the
effects we observed. We controlled for social conservatism throughout.
Method
Participants. We collected data from 875 U.S residents (M age= 25.09, SD =
8.59; 69.5% male) using the SocialSci online platform in December 2011. 454
participants identified as White, 253 as Black/African-American, 59 as Asian/Asian
American, and 21 as Latino/Hispanic American; 42 did not identify any racial/ethnic
group membership, and the remaining 43 individuals indicated a different category.
Given our small sample size of non-Black minorities, and because one of our central
questions of interest concerned perceptions of power differences between ethnic groups,
we focused our analyses on White and Black American participants (N = 707; M age=
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
19
25.17 SD = 7.86; 69.2% male), though we note that conclusions are unchanged when the
whole sample is used. 649 participants provided complete data on all focal variables.
Measures. Participants completed a battery of items related to social and political
attitudes collected for this and other purposes. We focus our analyses here on the
variables directly relevant to the research question at hand. Unless otherwise noted, all
items were assessed using a 1= ‘Strongly Disagree’ to 7= ‘Strongly Agree’ scale.
Independent Variables.
Social Dominance Orientation. Participants responded to the 16-item SDO-6
scale (Pratto, Sidanius, Stallworth, & Malle, 1994), as well as a variety of SDO items
intended for development of the SDO-7 scale (Ho et al., 2015). Although we center our
analyses on the SDO-6 scale items (α= .93), we note that we obtain the same results when
we use the items that ultimately formed the new SDO-7 scale (Ho et al., 2015)3.
Social Conservatism. Participants were asked: “In terms of social issues, how
would you describe your political attitudes and beliefs?” Responses were provided on a
1-7 scale (1= ‘Very Liberal’; 2 = ‘Liberal’; 3= ‘Slightly Liberal’; 4= ‘Middle of the
Road’; 5= Slightly Conservative; 6= ‘Conservative’; 7= ‘Very Conservative’).4
3 Note that we do not distinguish between different forms of anti-egalitarianism in the present work (see Ho et al., 2012; 2015), but rather are interested in how general support for group-based dominance and inequality relates to perceptions of social hierarchy. 4 We also had two items assessing economic conservatism and political party affiliation, respectively. Because of the potential overlap between anti-egalitarianism and each of economic conservatism and party affiliation (Jost et al., 2003; Kandler, Bleidorn, & Reimann, 2012), these were omitted from our analyses, and we instead focused on the most theoretically relevant contrast (i.e., to social conservatism). Nevertheless, patterns were highly similar when the economic conservatism and party affiliation items were included in a composite assessing conservatism (full details available from the first author on request). Results were also similar, across studies, when social conservatism was not included as a control variable in examining the effects of SDO (see section 3 of the Supplemental Materials for full details).
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
20
Social Class. Participants were asked: “How would you describe your family’s
social class position?” (1= ‘Poor; 2= ‘Working Class’; 3= ‘Middle Class’; 4= ‘Upper
Middle Class’; 5= ‘Upper Class’).
Dependent Variables.
Inter-Ethnic Power Differences. In order to assess perceptions of power
differences between the top of the ethnic group hierarchy and lower levels, we presented
participants with the following question: “How much power do each of the following
groups enjoy? Use a scale from 1 to 7, where 1 implies no power at all, and 7 implies
absolute power”. Participants rated the power of each of: Whites/Caucasian Americans,
Blacks/African Americans, Asians/Asian Americans, and Latino/Hispanic Americans.
We were primarily interested in perceived differences between the group at the top of
social hierarchy and those groups below. Because our sample was comprised of White
and Black Americans (and because we also account for ethnic group membership in our
analyses), we computed a difference score between ratings of White Americans (who, on
average, were rated as the ethnic group with the highest power; M = 5.84, SD = 1.55) and
Black Americans (who, on average, were rated as having significantly less power; M =
3.70, SD = 1.42, paired-samples t(693) = 27.63, p < .001).5 Higher scores thus indicate a
greater perceived power advantage for Whites (i.e., the high power group) relative to
Blacks (i.e., the low power group).6
5 To check for robustness, we also indexed perceived power differences as the difference between ratings of Whites’ power and the average perceived power for all other groups (α = .70). Conclusions are unchanged using this index. 6 Although we center our analyses on difference scores because of our central theoretical interest in perceptions of the gap between groups, the use of difference scores has certain disadvantages (e.g., Cronbach & Furby, 1970; Edwards, 1994; but see Thomas & Zumbo, 2012). Results examining ratings of the advantaged vs. disadvantaged group separately (for all relevant studies) can be found in the section 1 of the Supplemental Materials. Although the pattern differed somewhat across studies, we generally found
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
21
Inter-Sex Power Differences. We assessed inter-sex power differences similarly:
Participants rated the power of men (M = 5.80, SD = 1.48) and women (M = 4.43, SD =
1.33; paired samples t(696) =22.39, p < .001) using the same scale as above, and we
subtracted the rated power of women from the rated power of men, such that higher
scores indicated a greater perceived power advantage for men relative to women.
Inter-Class Power Differences. As above, participants rated the power of each of
the following groups: ‘Those born into a poor or working class family’, ‘Those born into
a comfortable or middle class family’, and ‘Those born into a wealthy or upper class
family’. From the rated power of those born into wealthy or upper class families (M =
6.00, SD = 1.40), we subtracted the average rated power of those born into comfortable or
middle class and poor or working class families (M = 3.41, SD = 1.15; paired samples
t(697) = 36.41, p < .001). Thus, higher scores indicated a greater perceived power
advantage for those born into rich families relative to others.
Social Welfare Support. Participants were asked about the extent to which they
supported each of the following policies: “Greater assistance to the poor”, “Reduced
public support for the homeless” (reverse-scored), and “Reduced benefits for the
unemployed” (reverse-scored) (α = .77).
Anti-Discrimination Policy Support. We assessed support for the implementation
of anti-discriminatory policy using 8 items. Example items include: “People have no
business trying to ensure racial integration in society” (reverse-scored)” and “Society
should do everything it can to help improve the economic condition of poor ethnic
minorities” (α = .91; see section 10 of the Supplemental Materials for full scale).
support for the idea that SDO is both negatively associated with perceived power of advantaged groups and positively associated with perceived power of disadvantaged groups.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
22
Affirmative Action Support. We assessed support for affirmative action by asking
individuals to indicate their support for each of the following policies: “Quotas, that is,
setting aside places for certain groups”, “Using memberships in certain groups as a tie-
breaker when applicants are equally qualified”, “Using group membership as one of
several considerations”, “Giving training to certain groups so they can compete equally”,
“Making a special effort to find qualified people from certain groups”, and “Giving
preference to members of certain groups who are less qualified than someone else.” (α =
.84).
Legacy Admission Policy Support. Participants were asked about their support
for university legacy admission policies using 4 items (see Ho et al., 2012; Kteily et al.,
2012). An example item is: “It is sometimes OK to consider family connections in
making admissions decisions” (α = .82).
New Hierarchy. Participants were presented with a measure of preferences
regarding the distribution of a hypothetical new university’s resources across schools (Ho
et al., 2012, 2015; Kteily et al., 2012). In order to assess preferences for a more
hierarchical distribution, we asked participants to indicate their responses to the following
statements: “I would prefer schools to be responsible for their own funding”, “I would
prefer the university to distribute resources equally rather than have each school fund
itself” (reverse-coded), “It would be unfair if schools had unequal budgets” (reverse-
coded), and “It would be fair for each school to get the budget it earns.” (α = .71).
Results and Discussion
We were interested in overall perceptions of power differences between the most
powerful groups and the rest of the groups in the hierarchy (across different types of
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
23
intergroup hierarchy). Thus, we first investigated whether perceptions of power
differences across the ethnic, gender, and class domains were inter-correlated. Indeed,
perceived ethnic power differences were positively correlated with perceived gender (r =
.55, p < .001) and class (r = .61, p < .001) power differences. Perceived gender and class
power differences were also highly correlated (r = .48, p < .001). A principal components
analysis using these three items generated a one-factor solution (eigenvalue= 2.09,
69.77% variance explained) and a reliable scale (α = .78), which we used as the central
variable (overall power differences) in our analyses.
We next examined whether SDO (M = 2.69, SD = 1.19) predicted this overall
index of power differences between high and low power groups. We observed that SDO
was significantly skewed in this sample (Skewness = .195, SE = .092, z = 2.12, p = .034).
Thus, we logged SDO for our analyses (results were consistent when SDO was not
logged).7 Consistent with our hypothesis, SDO was significantly associated with
perceptions of fewer overall power differences between groups (r = -.43, p < .001).
Because SDO is theorized to reflect support for hierarchy between groups in general, we
were primarily interested in the association between SDO and our overall power
differences index. Nevertheless, we also examined its association with perceived power
differences within each of the three domains examined. In all cases, SDO was
significantly associated with reporting a smaller power gap between advantaged and
disadvantaged groups (ethnic power differences: r = -.33, p < .001; gender power
differences: r = -.34, p < .001; class power differences: r = -.42, p < .001).
7 SDO was significantly skewed in some of our samples but not others (see section 2 of the Supplemental Materials for information about SDO’s distribution in each study). For the sake of consistency and comparability, we logged SDO in all studies. We note that conclusions are consistent throughout when SDO is not log-transformed.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
24
We subsequently considered whether reduced perceptions of power differences
between groups influenced support for egalitarian social policy. Specifically, we used
Hayes’ (2013) PROCESS macro (Model 4) to investigate a process model in which
overall perceptions of power differences mediated the effects of SDO on our various
measures of social policy. In order to test the unique effects of SDO, we controlled for
social conservatism’s effects— both on perceived power differences and on the outcome
measures— across analyses. We also controlled for participant gender (0= ‘Female’; 1=
‘Male’), social class, and racial minority status (0= ‘White’; 1= ‘Black’), in order to
isolate the effects of SDO from those of membership in high or low status groups
(conclusions were consistent when we did not control for group membership).
We assessed the effects for each of the various social policies separately.
Consistent with expectations, across 4 of the 5 dependent variables examined (i.e., all
except affirmative action support), SDO exerted significant negative indirect effects on
support for egalitarian social policy via its effects on decreasing perceptions of power
differences between groups (see Table 1). Thus, SDO was correlated with perceiving less
inequality between groups, which itself was associated with rejection of hierarchy-
attenuating (and endorsement of hierarchy-enhancing) social policies.8 We averaged
across the various policies examined in order to estimate an overall pattern for egalitarian
social policy (hierarchy-enhancing policies were reverse-scored).9 As with the individual
8 Further analyses with the affirmative action scale revealed that it was bi-dimensional, with one factor reflecting support for favoritism in hiring (e.g., use of quotas or tiebreakers), and another factor reflecting support for more restricted forms of affirmative action: training and outreach. Analyses conducted using the more restricted ‘outreach’ version of affirmative action support yielded the same significant indirect effect as for the remaining outcome measures. Unexpectedly, however, we observed the reverse pattern for the ‘favoritism’ affirmative action policies. This may due to low SDO individuals, who endorse egalitarianism, perceiving the favoritism element of these policies as unfair. 9 The scale incorporating support for each of the egalitarian policies had acceptable reliability (α = .58). Nevertheless, rejection of affirmative action was poorly correlated with the remaining measures (item-total
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
25
policies in isolation, perception of power differences was a significant mediator of SDO’s
effect on our composite measure of egalitarian social policy (see Table 1).
Finally, we assessed whether the indirect effect of SDO on rejecting egalitarian
social policy via perceived power differences was moderated by each of participants’ (a)
ethnic group membership, (b) gender, and (c) social class. We used Model 7 of Hayes’
(2013) PROCESS macro, which provides an index of moderated mediation. In fact, we
observed that the indirect effects were not moderated by ethnic group membership (index
of moderated mediation: -.00 [-.10, .10]) or by gender (index of moderated mediation: .01
[-.08, .10]): indirect effects were significant and equally strong among both Whites
(indirect effect: -.21 [-.29, -.14]) and Blacks (indirect effect: -.21 [-.31, -.14]), as well as
among men (indirect effect: -.20 [-.28, -.13]) and women (indirect effect: -.21 [-.30, -
.14]). Similarly, although we observed a significant (and unpredicted) tendency for the
indirect effects to be slightly stronger among individuals of lower social class (index of
moderated mediation: .05 [.01, .11]), the indirect effects were significant at all levels of
social class (at low social class: -.24 [-.33, -.17]; at mean social class: -.20 [-.27, -.14]; at
high social class: -.16 [-.24, -.09]).
In addition to the effects of SDO, we also observed that gender group membership
was associated with perception of power differences (with males, as expected, perceiving
less inequality). Unexpectedly, we observed that Whites perceived more inequality than
Blacks.10
correlation r = .04), and removing this item from the scale increased reliability (α =.67). Thus, reported results exclude affirmative action. However, we note that conclusions are unchanged when affirmative action support is included. 10 African Americans in this sample atypically expressed less egalitarian sentiments than White Americans across several indices. Nevertheless, as we note above, our hypothesized effect of SDO on perceived power differences held among both Whites and Blacks. Moreover, we observed this same relationship between
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
26
In sum, we observed strong support for our theoretical predictions. Perceptions of
power differences between groups across contexts were highly correlated (and formed a
unidimensional scale): those who thought that men enjoyed fewer advantages over
women tended to also think that Whites enjoyed fewer advantages over Blacks, and that
rich people enjoyed fewer advantages over other social classes. Moreover, across group
categories, and controlling for any effects of social conservatism and group membership,
these perceptions were significantly predicted by SDO. Thus, the more individuals were
motivated towards inequality between groups, the more likely they were to perceive the
relationship between dominant and subordinate groups as relatively equal, and thus to
reject egalitarian social policies (e.g., opposition to welfare and anti-discrimination
policies). Not only did this mechanism cut across a broad swathe of social categorizations
(i.e., among both men and women, minorities and Whites, and across social class
groups), it also held controlling for group membership, and was independent of the
effects of social conservatism.
Notably, we replicated our hypothesized patterns among a separate sample of
Whites (N = 498; see Supplementary Study 1). Specifically, SDO was uniquely
associated with perceiving fewer power differences between groups in the gender, racial,
and class domains; furthermore, this association had significant indirect effects on policy
support, and was not significantly moderated by participant gender or class.
Our theorizing proposes that SDO influences perceptions of power differences. At
the same time, cross-sectional analyses do not provide insights into causal relationships.
In Study 1b, we examined the relationship between SDO and perceptions of power SDO and perception of power differences (again, across race, gender, and class contexts) in a separate sample of Black Americans collected for other purposes using Qualtrics Panels (N = 620; r = -.33, p < .001).
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
27
differences between real-world groups longitudinally over a period of 4-6 weeks,
providing some insight into the validity of our proposed causal order. In line with our
theorizing, we hypothesized that SDO at time 1 would predict perceived power
differences at time 2, controlling for these perceptions at time 1.
Study 1b
We collected data from a sample of 355 residents of the U.S. (M age = 34.10, SD
= 11.87; 51.1% male) using Amazon’s mTurk platform, and followed up with them 4-6
weeks later. 252 participants (71.0%) completed the second wave of the survey, of whom
244 provided data on all focal variables.11
Measures.
Social Dominance Orientation was assessed using the 16-item SDO-7 scale (Ho
et al., 2015; Wave 1: α= .93; Wave 2: α= .94).
Overall Power Differences was assessed as in Study 1a, using participants’
ratings of inter-ethnic, inter-sex, and inter-class power differences, which again formed a
reliable (Wave 1: α = .73; Wave 2: α =.70) and unidimensional (Wave 1: eigenvalue =
1.94, 64.7% variance explained; Wave 2: eigenvalue = 1.90, 63.32% variance explained)
scale. Ratings of power for each group were provided on a 0 (‘No power at all’) to 100
(‘Absolute power’) scale. Because we were not focusing our assessment on White and
Black participants here (as we had in Study 1a), we expanded our measure of inter-ethnic
power differences. Specifically, we assessed this variable by examining the gap between
ratings of the perceived power of Whites and the average rated power of all other groups
(i.e, Black, Hispanic, and Asian Americans; Wave 1: α = .85; Wave 2: α = .87). On
11 Participants recruited via mTurk were restricted from participating in more than one of the studies using mTurk reported in this paper. Due to a programming error, participant race was not recorded in Study 1b.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
28
average, Whites (Wave 1: M = 76.68, SD = 19.57; Wave 2: M = 77.79, SD = 17.45) were
rated as having significantly more power than the remaining ethnic groups (Wave 1: M =
42.49, SD = 21.09; paired samples t (350) = 24.42, p < .001; Wave 2: M = 42.00, SD =
20.58; paired samples t (249) = 22.75, p < .001).
Results and Discussion
Participation attrition was not significantly associated with Wave 1 levels of
either SDO (F (1, 347) = 1.62, p = .20) or of perceived power differences (F < 1).
We were centrally interested in examining whether SDO predicted perceptions of
power differences over time. As hypothesized, SDO in Wave 1 predicted perceiving
fewer power differences between groups in Wave 2, controlling for perceived power
differences in Wave 1, b = -1.63, β = -.10, p = .037, 95% CI [-7.17, -.22]. On the other
hand, we observed that perceptions of power differences in Wave 1 were not significantly
associated with SDO levels in Wave 2, controlling for SDO levels in Wave 1, b = -.00, β
= -.04, p = .24, 95% CI [-.003, .001]. Moreover, using Hayes’ (2013) PROCESS macro
(Model 1), we observed that the relationship between SDO in Wave 1 and perceived
power differences in Wave 2 (controlling for this variable in Wave 1) was not
significantly moderated by either of gender (b = -2.68, p = .41, 95% CI [-9.05, 3.69]) or
social class (b = -1.87, p = .37, 95% CI [-5.94, 2.20]).
In sum, we obtained evidence consistent with our argument that SDO influences
downstream perceptions of inequality between groups, whereas we did not find any
evidence consistent with the reverse causal pathway. Although this study supports our
theoretical perspective, it should be noted that we examined this question over a short
time period (providing less opportunity for change in our constructs). A more
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
29
comprehensive examination could assess these patterns over greater periods of time, and
could consider assessing developmental time-points (e.g., college years; Alwin, Cohen, &
Newcomb, 1991) during which sociopolitical attitudes are known to be heavily shaped
(we discuss causal order further in the General Discussion).
The patterns we observed across our first two studies (and Supplemental Study 1)
are consistent with the idea that variation in individuals’ equality motives leads them to
perceive power differentials between groups differently, and suggest that this relationship
does not differ significantly as a function of membership in advantaged versus
disadvantaged groups. At the same time, it is not clear from these studies whether this
perceptual bias arises from high vs. low SDO individuals being exposed to fundamentally
different information about inequality (e.g., as a result of living in different
neighborhoods or choosing different jobs), or from processing the same underlying
information differently (e.g., by paying more attention to and more heavily weighting
worldview-consistent aspects of the information). In order to determine whether the
association between SDO and perceived power differences might reflect motivated
differences in the way individuals process inequality-relevant information, we need to
ensure that individuals are being exposed to the same information.
In Study 2 (and Supplementary Study 2), we sought to examine whether the
association between SDO and inequality perception would hold when participants rated
power differences between groups in a fictional asymmetric intergroup conflict. Given
that all participants read the same text, we are able to rule out any role here of differential
exposure to information relating to power differences. Moreover, by moving beyond
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
30
judgments of real-world power differences, we were able examine the generality of the
phenomenon under investigation.
Study 2
Method
We collected data from 164 U.S. residents (M age = 35.33, SD = 12.65; 52.6%
male; 116 White/European American; 12 Asian/Asian American; 11 Latino/Hispanic
American; 9 Black/African American; 4 biracial/mixed race; 1 Other; 11 missing) on
Amazon’s mTurk platform, of whom 153 provided data on all focal variables. We asked
participants to read about a fictitious asymmetric conflict taking place between two
groups, a scenario modeled on one previously determined to reflect an asymmetric
intergroup conflict (Kteily et al., 2013, Study 3). Specifically, participants were first told
that they would be reading about a conflict between two groups in a country called
‘Raga’, and that they would be asked to represent one of the two groups involved in the
conflict. Participants were randomly assigned to take on the role of either ‘Group A’ or
‘Group B’, pre-determined to be perceived as high vs. low in power on average,
respectively (Kteily et al., 2013). Subsequently, all participants read about the history of
the conflict: specifically, they were told that Group A and Group B both made claims to
all the land in Raga, with tensions resulting in a war between the two sides. Group A had
won the war, taking control of Raga and the vast majority of its resources. Group B were
said to feel wronged by their situation, and to have been engaging in both violent and
nonviolent resistance efforts in order to improve their situation, efforts which had scored
some successes. Previous research had shown that assignment into the high vs. low power
group in this scenario had psychological consequences (i.e., differentially influencing the
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
31
conditions under which individuals were willing to negotiate with the other side; Kteily et
al., 2013).
Measures
Independent Variables.
Social Dominance Orientation. This construct was assessed as in Study 1a (α=
.92).
Social Conservatism. This construct was assessed as in Study 1a.
Dependent Variables.
Power Differences. We assessed perceptions of power differences as we had
assessed it in Study 1a. Thus, we asked participants: “How much power do each of the
following groups enjoy?” Participants were asked to provide their responses for each of
Groups A and B, using a 0 (‘No power at all’) to 100 (‘Absolute power’). We subtracted
ratings of the low power group (Group B) from ratings of the high power group (Group
A), with greater scores reflecting a larger advantage for the high power group.
Support for Egalitarian Social Policy. We assessed participants’ support for
egalitarian policies or actions intended to challenge the hierarchical status quo between
the two sides by asking participants to indicate their support for each of the following 4
statements: “I support affirmative action programs for members of Group B in Raga”,
“International forces should enter Raga and ensure the equal treatment of members of
Group A and Group B”, “I would support the establishment of an international
commission to investigate any injustices committed by Group A against Group B”, and “I
think the international community should place a great deal of pressure on both groups
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
32
until they agree a resolution.” Responses were provided on a 1 (‘Strongly disagree’) to 7
(‘Strongly agree’) scale (α = .83).
Results and Discussion
Group A (M= 83.98, SD = 14.37) was, on average, rated as significantly more
powerful than Group B (M = 19.33, SD =16.90), paired-samples t(155) = 27.02, p < .001.
We examined our hypotheses using Hayes’ (2013) PROCESS macro (Model 4),
controlling for social conservatism and membership in the high vs. low power group.
There was a nonsignificant trend such that being a member of the high power group
(Group A) was associated with a reduced perception of power advantage (b = -7.78, β = -
.13, p = .11, 95% CI: [-17.43, 1.58]). Social conservatism was not associated with power
perceptions (b = -.11, β = -.01, p = .94, 95% CI: [-2.99, 2.41]). On the other hand, and as
in previous studies, SDO was associated with perceiving less of a power advantage for
the dominant group (b = -18.88, β = -.27, p = .002, 95% CI: [-31.15, -7.67]). Moreover,
perceptions of a bigger advantage for the powerful group were associated with more
support for egalitarian social policies (b = .01, β = .24, p = .004, 95% CI: [.005, .02]).
The hypothesized indirect effect was significant (indirect effect: -.22 [-.43, -.09]).
Notably, and consistent with the idea that assignment into group was psychologically
meaningful, we further observed that beyond SDO, participants assigned to the high
power role were significantly less supportive of egalitarian policies in the Raga context (b
= -.51, β = .18, p = .02, 95% CI [-.93, -.08].
We next assessed whether the pattern associated with SDO were moderated by
group membership, examining whether the indirect effects differed among participants
assigned to play the role of the high power (Group A) vs. low power (Group B) group.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
33
Despite the fact that those assigned to play the high power role were less supportive of
egalitarian policies, the indirect effects of SDO on support for egalitarian actions via
perceptions of power differences were not significantly moderated by group membership
(index of moderated mediation: -.19 [-.53, .05]). Indeed, the indirect effects were
significant for both the low power (indirect effect: -.15 [-.38, -.009]) and high power
(indirect effect: -.34 [-.66, -.13]) groups.
Thus, we obtained results in a fictional intergroup context that were consistent
with those observed among real-world groups. The greater individuals’ motivational
orientation towards inequality between groups, the smaller the power differential they
perceived between the advantaged and the disadvantaged group. This was true controlling
for individuals’ social conservatism, and occurred even though all participants in this
study were exposed to exactly the same information about the intergroup relationship.
Consistent with prior studies, perceiving a smaller advantage for the high power group
was associated with less support for policies intended to revise the hierarchical status
quo. Notably, we replicated this indirect association between SDO and policy support via
perceived power differences in a separate sample using a different fictional intergroup
context (see Supplemental Study 2).
Furthermore, we observed similar patterns here across membership in high vs.
low power groups: whether individuals were assigned to the high or low power group
role, lower (vs. higher) levels of SDO led to a greater perceived power differential
between groups, which was associated with more support for policies that would revise
an unequal status quo.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
34
In Studies 3a-3c, we examined whether the relationship between equality motives
and perceptions of hierarchy extended even further, to basic and abstract visual
representations of power differences. As in Study 2 (and Supplemental Study 2), we
presented all participants with the same information (here, visual), again examining
differential processing of the same underlying information about hierarchy.
Study 3a
Method
We collected data from 171 U.S. residents (M age = 31.90, SD = 10.49; 57.1%
male; 122 White; 13 Asian/Asian American; 5 Black/African-American 4
Latino/Hispanic American; 2 Jewish; 7 Biracial/mixed race; 1 Other; 17 missing) on
Amazon’s mTurk platform, of whom 156 provided data for all focal variables. We asked
participants to look at a randomized series of five images, one at a time, each of which
they were told represented a hypothetical society. Each image consisted of a ladder with
ten rungs, with a number of bags of money sitting adjacent to each rung (see Figure 1; see
also Goodman et al., 2001; Sheehy-Skeffington, 2015). Participants were told that each
rung represented 10% of the people in the fictitious society, and that the bags of money
represented the wealth of that population decile. Power was operationalized as wealth in
this study consistent with perspectives suggesting that access to resources is an important
basis of power (Dubois, Rucker, & Galinsky, 2015; Emerson, 1962; Keltner, Gruenfeld,
& Anderson, 2003). We used multiple images so that we could construct a reliable and
valid measure of hierarchy perceptions (Judd, Westfall, & Kenny, 2012), systematically
varying each image so that it depicted a different level of inequality in power (resulting in
a ratio of the wealth held by the top decile to the bottom decile ranging from 1.67 to 14).
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
35
While viewing each image, participants responded to a set of questions about that
image, before moving onto the next. Once five images had been shown, participants were
presented with measures of our SDO and social conservatism (in randomized order),
followed by a sociodemographic questionnaire.
Measures
Power Differences. In order to assess individuals’ perceptions of the degree of the
hierarchy between those at the top and at the bottom of the fictitious societies, we asked
participants to respond on a 7-point Likert scale (1 = Strongly disagree to 7 = Strongly
agree) to the following five questions, as they viewed each image: “The distribution of
money in this society is very unequal”, “There is not a big difference in the distribution
of money from the bottom to the top of this society” (reverse-coded), “Resources in this
society are fairly equally spread” (reverse-coded), and “The hierarchy that exists in this
society is very steep”, and “This society is not really a hierarchical one” (reverse-coded).
We computed our index of perceived power differences in two steps. First, we created a
composite using the mean score across all items assessing perceptions of power
differences for each individual image (αs= .77 to .89). Next, we computed a composite
from the mean scores across all five images (α = .65).12 As before, higher scores indicate
a greater perception of a power differential across the hierarchy.
Support for Egalitarian Social Policy. We assessed participants’ support for
policies or actions intended to revise the existing hierarchical status quo by asking them
to use the same 7-point scale (1 = Strongly disagree to 7 = Strongly agree) to respond to 12 Due to the relatively low Cronbach’s α value for this composite, we conducted an analysis of inter-item correlations, revealing that the removal of the mean score from the first (i.e. most equal) ladder image yielded a more reliable 4-item measure (α = .76). We conducted all reported analyses with the 4-item version of this scale. However, analyzing results using the 5-item version yielded equivalent results on all tests.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
36
the following five statements for each image: “Those at the top of this society should
share more of their wealth with those at the bottom”, “There is a need to redistribute
wealth from the top to the bottom in this society”, “There is no need to change the
income distribution in this society.” (reverse-coded), “There is a need to flatten the
hierarchy in this society.”, and “Support should be given for those at the bottom of this
society to receive more money.” As with perceived power differences, we created a
composite measure of support for egalitarian social policy by first computing means
across all items for each of the images (αs =.90 to .93), and then computing a mean
across all images (using all five images, α = .88; using the four most reliable images, α =
.93).
SDO. Participants completed the 16-item SDO-6 measure (Pratto et al., 1994), as
in Study 1a (α= .85).
Social Conservatism was assessed as in previous studies.
Results and Discussion
Replicating our previous findings, controlling for social conservatism (b = -.04, β
= -.09, p = .34, 95% CI: [-.13, .04]), SDO significantly predicted perception of power
differences, this time using abstract visual representations of hierarchy, b =-1.10, β = -
.42, p < .001, 95% CI: [-1.54, -.65]. Moreover, as in previous studies, perceiving greater
power discrepancies was significantly associated with more support for egalitarian
changes to the status quo (b = .63, β = .40, p < .001, 95% CI: [.38, .86]). We used Hayes’
(2013) PROCESS macro (Model 4) to investigate whether SDO exerted a significant
indirect effect on support for egalitarian social policy via power perceptions, controlling
for social conservatism. Indeed, the hypothesized indirect effect was significant (indirect
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
37
effect: -.69 [-1.14, -.29]). In sum, the more individuals were motivated to maintain
hierarchy between groups in society, the less they perceived abstract depictions of
unequal societies as hierarchical, and the more they rejected policies intended to equalize
resource distribution between those at the top and those at the bottom.
We thus replicated the link between SDO, perceptions of hierarchy, and support
for egalitarian social policy in an artificial context that was presented in the abstract form
of a simple set of images. This provides further evidence that the relationship between
SDO and perceptions of inequality extends even to contexts in which individuals are
exposed to identical underlying information. Furthermore, it extends the findings of
Study 2, which used a detailed description of conflict between two groups, to suggest that
social dominance motives can shape the perception of inequality depicted even pictorially
in ways that contribute to the protection of dominant group advantages. Once again, this
link held controlling for the effects of social conservatism.
In study 3b, we created a novel set of images of intergroup hierarchy that centered
on social (rather than resource-based) power, thus allowing us to test the generalizability
and replicability of the findings of Study 3a across abstract operationalizations of power.
Study 3b
Method
We collected data from 240 U.S. residents (M age = 34.60, SD = 11.22; 51.3%
male; 170 Caucasian/White; 22 Asian/Asian American; 15 Black/African American; 6
Latino/Hispanic; 5 Jewish; 6 biracial/mixed race; 16 missing) on Amazon’s mTurk
platform, of whom 227 provided data on all focal variables. Similarly to Study 3a, we
asked participants to look at a series of five images, and presented them with questions
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
38
regarding the degree of hierarchy depicted in the images, and the need for challenge of
the status quo. Consistent with recent calls for the diversification of stimuli (Judd et al.,
2012), and in order to ensure that the findings of study 3a weren’t dependent on the type
of image (ladder) or of resource inequality (money), we designed a novel set of
hierarchies modeled on organizational pyramids. Each image consisted of a pyramid with
multiple layers, and different numbers of stick figures at each layer (see Figure 2).
Participants were asked to imagine that each pyramid represented a fictitious
organization, that each layer represented a level of power in that organization, and that
the stick figures represented the number of people at that level. We again used multiple
images so that we could construct a reliable and valid measure of hierarchy perceptions,
systematically varying each image so that it depicted a different level of inequality in
power (resulting in a number of layers that ranged from 4 to 11, and a ratio of the number
of people in the top two layers to the bottom two layers that varied from 0.22 to 0.04).
Our theorizing suggests that SDO motives influence the perception of hierarchy
between groups (rather than the reverse). In the previous studies, we assessed SDO prior
to perceived hierarchy, in line with our proposed causal order. In Study 3a, however, we
measured SDO after perceptions of hierarchy, because we reasoned that examining
perceptions of hierarchy without directly making participants aware of our interest in
their levels of SDO would represent a particularly conservative test. In order to directly
examine whether order of measurement of SDO matters, we counterbalanced the order of
SDO (and social conservatism) in Study 3b, assessing it either prior to, or after, ratings of
perceived hierarchy.
Measures
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
39
Power Differences. In order to assess individuals’ perceptions of how much
hierarchy existed between groups of individuals at the top and at the bottom of the
fictitious organizations we presented, we asked participants to respond on a 7-point
Likert scale (1 = Strongly disagree to 7 = Strongly agree) to the following five questions,
as they viewed each image: “The distribution of power in this organization is very
unequal”, “There is not a big difference in the distribution of power from the bottom to
the top of this organization” (reverse-coded), “This organization is fairly ‘flat’” (reverse-
coded), and “The hierarchy that exists in this organization is very steep”, and “This
organization is not really a hierarchical one” (reverse-coded). As in Study 3a, we
computed our index of power differences by first creating a composite using the mean
score across all items for each image (αs ranged from .80 to .88), and then computing a
composite as the mean score across all images (α= .75). As in previous studies, higher
scores indicated a greater perception of advantage for groups with high power (top
organizational layers) relative to those with low power (bottom organizational layers).
Support for Egalitarian Social Policy. We assessed participants’ support for
policies intended to reduce inequality by asking them to respond to the following 5
statements for each image: “Those at the top of this organization should share more of
their power with those at the bottom”, “There is a need to redistribute power from the top
to the bottom in this organization”, “There is no need to change the power distribution in
this organization” (reverse-coded), “There is a need to flatten the hierarchy in this
organization”, and “Support should be given for those at the bottom of this organization
to move higher.” Participants again indicated their perceptions using a 7-point scale (1 =
Strongly disagree to 7 = Strongly agree). As with power perceptions, we created a
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
40
composite measure of support for egalitarian social policy by first computing means
across all items for each images (αs ranged from .85 to .89), and then computing a mean
across all images (α= .92).
Social Dominance Orientation. This construct was assessed as in Study 3a
(α=.95).
Social Conservatism was assessed as in prior studies.
Results and Discussion
We observed that order of SDO measurement had no effects on mean levels of
SDO (F < 1), on mean perception of power differences (b = -.09, β = -.05, p = .37, 95%
CI [-.27, .10]), or on the relationship between SDO and perception of power differences
(b = .28, p = .12, 95% CI: -.08, .65]. Thus, we collapsed across this factor for all
remaining analyses; analyzing results in each of the experimental conditions separately
yielded conclusions consistent with those reported below.
As in Study 3a, we found a significant relationship between SDO and the
perception of power differences across the visually-depicted social system (b =-.59, β = -
.39, p < .001, 95% CI: [-.81, -.39]), controlling for social conservatism (b = -.04, β = -.10,
p = .17, 95% CI: [-.10, .02]). Using Hayes’ (2013) PROCESS macro (Model 4), we again
observed that the hypothesized indirect effect from SDO to support for egalitarian
changes to the status quo via perception of power differences was significant (indirect
effect: -.27 [-.42, -.15]): thus, SDO was associated with reduced support for redistributive
policy in the depicted hierarchical organizations through its link with decreased
perceptions of a power advantage for groups at the top relative to those at the bottom. We
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
41
further replicated these patterns in a separate sample (N = 170; see Supplementary Study
3).
Beyond suggesting that the results of Study 3a are unlikely to have depended on
the order of assessment of SDO, these findings extend the results of Study 3a to a second
abstract visual representation of power differences. Thus, the results of Study 3b provide
further evidence for the generality of the relationship we observe: Whether the social
hierarchy involves real or experimental groups, whether individuals themselves belong to
advantaged or disadvantaged groups, and whether the intergroup relationship is
represented using text descriptions or abstract images, individuals differ in their
perception of the power gap between those at the top and those at the bottom as a
function of their motivation toward hierarchy between groups in society. Moreover, those
individuals perceiving more hierarchy between groups indicate greater support for
policies intended to equalize conditions than do those who see power relations as
relatively equal.
In Study 3c, we sought to examine whether the patterns we observed using SDO
would generalize to other measures of (anti-)egalitarianism.
Study 3c
Method
We collected data from a sample of 236 residents of the U.S. on Amazon’s mTurk
(M age = 34.29, SD = 11.38; 55.2% male; 171 Caucasian/White; 16 Asian/Asian
American; 17 Black/African American; 13 Latino/Hispanic; 1 Jewish; 1 biracial/mixed
race; 1 Native American; 1 Middle Eastern; 15 missing), of whom 225 provided data on
all focal variables.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
42
Measures. We used the same stimuli and procedure as in Study 3b, but used
different measures of egalitarianism and traditionalism (in place of SDO and social
conservatism, respectively). Egalitarianism and traditionalism measures appeared prior to
ratings of perceived power differences.
For our assessment of egalitarianism and traditionalism, we primarily drew on
classic scales of political attitudes (see Table 3-3 in Knight, 1999). Some of these scales
embed the assumption that (economic) conservatism directly reflects anti-egalitarianism,
whereas others treat these as theoretically separate constructs. This parallels the fact that
some theoretical perspectives assume conceptual overlap between economic
conservatism and acceptance of inequality (e.g., Duckitt, 2001; Jost et al., 2009; Kandler,
Bleidorn, & Reimann, 2012), whereas others emphasize the possibility that individuals
embrace economic conservatism not because they are inegalitarian, but for ‘principled’
reasons (e.g., because they believe that economically conservative policies like low
taxation rates are more efficient and provide incentives and opportunities to all for
individual advancement; e.g., Tetlock & Mitchell, 1993).
Because we were primarily interested in (anti-) egalitarianism per se, we centered
our analyses on those items directly assessing attitudes about equality, and include
analyses using items assessing economic conservatism in the Supplemental Materials (see
section 7).
Anti-egalitarianism. We assessed anti-egalitarianism using two separate
measures. For our first metric, we adapted three items from a scale developed by
McCloskey and Bann (1979, as cited in Knight, 1999; sample item: “Efforts to make
everyone as equal as possible should be decreased”; α= .72; see section 10 of
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
43
Supplemental Materials for full scale). For our second metric of egalitarianism, we used a
five-item scale of economic egalitarianism used by Hatemi et al. (2014) (sample item: “If
wealth were more equal in this country, we would have fewer problems” (reverse-coded);
α= .91).
Traditionalism. We assessed traditionalism using two separate measures. First,
we adapted seven items from the same McCloskey and Bann (1979, as cited in Knight,
1999) scale (sample item: “Trying to make sweeping reforms in a society as complicated
as ours is usually much too risky”; α= .76; see section 10 of Supplemental Materials for
full scale). Our second metric of traditionalism adapted five items assessing ‘traditional
morality’ from Ray’s (1983, as cited in Knight, 1999) scale of conservatism (α= .67;
sample item: “Law and order is more important than letting every kook have their say”;
see section 10 of Supplemental Materials for full scale).
Power Differences. This construct was assessed using ratings of the same abstract
depictions of fictitious organizations as in Study 3b (α= .80).
Support for Egalitarian Social Policy. This construct was assessed as Study 3b (α
= .94).
Results and Discussion
We examined whether anti-egalitarianism was associated with perceptions of the
degree of hierarchy between those at the top and at the bottom of the fictitious societies,
controlling for traditionalism. Using the measures of anti-egalitarianism and
traditionalism adapted from McCloskey and Bann (1979), we observed that anti-
egalitarianism was significantly negatively associated with perceived hierarchy (b = -.16,
β = -.27, p = .001, 95% CI [-.25, -.07]). On the other hand, the negative association for
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
44
traditionalism was marginally significant (b = -.12, β = -.14, p = .06, 95% CI [-.25, .01]).
We obtained similar results when we instead used Hatemi et al.’s (2014) anti-
egalitarianism measure and Ray’s (1983) traditionalism measure: Specifically, anti-
egalitarianism was uniquely associated with perceiving less hierarchy (b = -.15, β = -.27,
p < .001, 95% CI [-.21, -.08]); here, traditionalism was significantly related too (b = -.19,
β = -.25, p < .001, 95% CI [-.29, -.09]).
Furthermore, as in earlier studies, we observed a significant indirect effect from
generalized anti-egalitarianism beliefs to support for context-specific egalitarian social
policies via perceptions of hierarchy between the relevant groups: Using Hayes’ (2013)
PROCESS macro (Model 4), we observed that this indirect effect was significant for both
metrics of anti-egalitarianism (as assessed by McCloskey and Bann (1979, as cited in
Knight, 1999): indirect effect = -.06 [-.12, -.02]; as assessed by Hatemi et al. (2014):
indirect effect = -.05 [-.10, -.02]). Notably, results were comparable irrespective of which
combination of anti-egalitarianism and traditionalism measures we used.
In sum, the results of Study 3c are highly consistent with the idea that individuals’
equality-motives are associated their perception of hierarchy beyond traditionalism.
Moreover, the results of this study suggest that the patterns we observe are not restricted
to one measure of anti-egalitarianism (i.e., SDO), but rather generalize broadly to
measures reflecting beliefs about the desirability of social equality. Interestingly, we
further observed that results using economic conservatism (adapted from McCloskey and
Bann, 1979, as cited in Knight, 1999) closely paralleled those obtained using anti-
egalitarianism (see section 7 of the Supplemental Materials), consistent with theorizing
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
45
suggesting that anti-egalitarianism is a basis of political conservatism (e.g., Jost et al.,
2009).
In our next study, we sought to arbitrate between two possibilities that could account
for the relationship between anti-egalitarianism and perceptions of hierarchy. It is
possible that, regardless of their motivational orientations towards inequality, individuals
perceive the same extent of hierarchy, but consciously report it differently for strategic
reasons. Thus, even if egalitarians and inegalitarians perceive the abstract depictions of
hierarchy that we show them in exactly the same way, egalitarians individuals might
consciously over-report hierarchy and/or inegalitarian individuals might consciously
under-report it, in order to justify their desire vs. distaste (respectively) for redistributive
policies. Alternatively, consistent with research on unconscious motivated information
processing biases (e.g., Kahan et al., 2012), we reasoned that individuals may come to
actually see the reality that conforms with their equality motives, thus perceiving the
same abstract depictions of hierarchy differently.
To begin distinguishing between these possibilities, we experimentally
manipulated whether or not participants had an incentive to report their perception of
hierarchy as they honestly saw it. We reasoned that if this incentive changed the
relationship between ideology and power perceptions, then participants might well be
seeing one thing but consciously and strategically reporting another in order to ward off
demands for greater equality. On the other hand, if the incentive for honest reporting had
no effect on hierarchy perceptions, this would be more consistent with the idea that their
social dominance motives were unconsciously influencing how they truly perceived
hierarchy.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
46
Study 4
Method
We collected data from 335 U.S. residents (M age = 33.99, SD = 10.99; 50.5%
male; 246 Caucasian/White American; 25 Black/African American; 25 Asian/Asian
American; 18 Latino/Hispanic American; 4 Native American; 2 Jewish; 1 Middle
Eastern; 11 biracial/mixed race; 1 Other; 2 missing) on Amazon’s mTurk platform, of
whom 328 provided data on all focal variables. The protocol was identical to that in
Study 3b, with the following exception: We randomly assigned participants to receive (or
not receive) an instruction intended to incentivize honest reporting. In the incentivized
condition, participants read the following instruction, adapted from Waytz, Young, and
Ginges (2014), who showed that it successfully reduced a form of biased perception (i.e.,
the ‘motive attribution asymmetry’): “You will see a series of 5 images, and will be
asked questions about each one. The participant whose responses are most accurate
across images will receive a $12 bonus on mTurk.” In a separate pre-test we conduced on
mTurk (n = 58), we observed that this manipulation was successful in influencing
participants’ approach to the task: specifically, we found that participants in the
experimental (vs. control) condition were significantly more likely to report being
motivated to respond with accuracy, F (1, 56 = 19.66, p < .001, partial η2 = .26).13
Social Dominance Orientation (α = .95) and Social Conservatism were assessed,
followed by Power Differences (α= .73). These constructs were assessed as in Study 3b.
We did not assess support for egalitarian policy in this study because our instructions to
13 Motivation to respond with accuracy was assessed using 9 items, including “I would not be very disappointed if I was less accurate than others on this task” (reverse-coded) and “I would be absolutely delighted if I was the most accurate person on this task” (α = .79).
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
47
participants in the experimental condition focused on accuracy in perceptions of the
images (rather than ratings for the appropriate policies).
Results and Discussion
Using Hayes’ (2013) PROCESS macro (Model 4), we assessed whether
experimental condition moderated the association between SDO and perceptions of
power differences. Consistent with prior studies, SDO was negatively associated with
perceptions of power differences (b= -.39, β = -.26, p < .001, 95% CI: [-.59, -.21]),
controlling for social conservatism (b= -.00, β = -.00, p = .98, 95% CI: [-.06, .05]).
Experimental condition (i.e., control vs. incentivization) had no main effect on perceived
power (b= .02, β = .02, p = .78, 95% CI: [-.14, .20]). Moreover, it did not moderate the
relationship between SDO and perceived power differences (b = .07, p = .67, 95% CI: [-
.25, .39]). Indeed, controlling for social conservatism, SDO was significantly associated
with reduced power perceptions in both the control (b= -.43, p < .001, 95% CI: [-.67, -
.19]) and incentivized (b = -.36, p = .002, 95% CI: [-.59, -.13]) conditions. Examining the
moderation from the other perspective, the incentive had no effect on either those lower
(b = -.01, p = .92, 95 % CI: [-.25, .22]) or higher in SDO (b = .06, p = .62, 95% CI: [-.18,
.29]).
In sum, incentivizing participants to honestly report their perception of power
differences in an abstract hierarchy had no influence on the relationship between higher
SDO and perception of less inequality (see also Supplemental Study 4, which obtained
highly consistent results using a similar paradigm). The fact that a strong incentive
manipulation which we showed had effects on individuals’ motivations to respond
accurately had no discernible influence on the SDO-power perception relationship
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
48
suggests that individuals’ underlying orientations towards equality may influence their
actual perception of hierarchy, and not just their reporting of it. It is important to state
that the lack of moderation cannot readily be explained by lack of statistical power, given
that we purposely ensured that each of the experimental and control conditions had large
sample size (over 150 individuals per condition), and found the hypothesized
relationships in both conditions.
In our final study, we sought to develop a paradigm that would allow us to
examine whether our effects extended to biased memory. If participants perceive
hierarchy differently as a function of their equality motives, they might also be expected
to remember it in a biased manner. In designing this study, we also sought to develop an
objective metric with which to assess participants’ cognitions about hierarchy as a
function of their motivation toward hierarchy. One important benefit of developing an
objective metric of hierarchy perception is its ability to shed light on whether our effects
result from high SDO individuals under-perceiving hierarchy, low SDO individuals over-
perceiving it, or both (see Chambers et al., 2014, 2015). Because our previous studies
were based on individuals’ subjective ratings of hierarchy, it was not possible to clearly
distinguish which of these patterns was responsible for the differences in hierarchy-
perception.
Study 5
In Study 5, we sought to examine whether individuals might show evidence of
biased memory for hierarchy as a function of their underlying orientation toward
hierarchy. Specifically, we asked whether individuals higher in SDO might misremember
having seen less hierarchy than they actually had. Simultaneously, we wanted to examine
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
49
the possibility that individuals lower in SDO might misremember having seen more
hierarchy than they had. In order to do this, we presented individuals with a series of
images depicting hierarchical organizations, and subsequently spot-checked their
memory by asking them to identify which images they had seen from an array of images,
which included the real image as well as more and less hierarchical versions. Thus, we
were able in this study to tap into objective accuracy in the perception of hierarchy, and
to assess it using a measure other than the rating of hierarchy: the identification of which
hierarchy individuals believe they had previously seen.
Method
Because we had developed this as a new paradigm, we had no objective metric by
which to determine an appropriate sample size, though we suspected that effects, if any,
would be small. We originally conducted this study among 226 U.S. residents on
Amazon’s mTurk. After analyzing the suggestive results of the first study, we decided to
conduct an exact replication with a new sample of 351 U.S. residents on mTurk. Because
the two studies were conducted in an identical fashion among the same population, and
yielded extremely similar results, we decided to combine these two datasets here for
brevity, and report the results of this pooled analysis (the full results of the original test
and the exact replication are reported separately in section 8 of the Supplemental
Materials). The final sample thus consisted of 577 participants (M age= 33.29, SD =
11.62, 57% female; 402 Caucasian/White American; 37 Black/African American; 34
Latino/Hispanic American; 29 Asian/Asian American; 8 Native American; 7 Jewish; 3
Middle Eastern; 20 biracial/mixed race; 6 other; 31 missing), of whom 539 provided data
for all focal constructs.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
50
As stimuli in this study, we constructed 4 sets of images depicting hierarchical
organizations, similar to those used in Studies 3b and 4. Each set had a different color
background, and contained 5 images of organizational pyramids, with varying levels of
hierarchicalness. We selected the images based on a pretest, in which we asked a separate
sample of participants (n = 208) to use a shortened version of the hierarchy-rating scale
from Studies 3b and 4 (consisting of 3 of the 5 items) to rate the power differences in 5
sets of 7 images, designed to have differing objective levels of hierarchy.14 Using this
data, we selected 4 image sets for the current study, each containing 5 images— one
‘focal’ image and four ‘distractor’ images— based on an optimization of the difference in
subjective ratings of degree of power differences across all 5 images.15
Participants were told that we were conducting a study about “perception of
images”, for which they would be shown and asked to make judgments about a series of
4 images. Participants were first shown each of the 4 focal images in randomized order,
as part of the “Images task”. As in Study 3b, participants rated the degree of power
differential they perceived in each of the ‘organizations’ depicted in the images. After
they had completed their ratings, participants were surprised with a “Memory task”.
Specifically, we told participants that they would be presented with 4 sets of images in an
14 We used three indicators to objectively vary and track the degree of hierarchy in each organization: the number of rows in that organization (with more rows meaning greater degree of hierarchy or difference in power), the height of the triangle depicting the organization (taller triangles conveying greater hierarchy), and the ratio of the number of employees in the top-two ‘leadership’ rows to the bottom two 'subordinate' rows (a smaller ratio indicating greater hierarchy). 15 Specifically, the primary criterion was that the difference between the average of the subjectively-rated hierarchicalness between the two least hierarchical images and the hierarchicalness of the middle, focal image was equal (to an accuracy of 0.1 on a 9-point scale) to the parallel difference in subjectively-rated hierarchicalness between the focal image and the average hierarchicalness of the two most hierarchical images. Secondarily, our selection also kept in mind a desire for (1) equivalence of the difference between the least hierarchical image and the second-to-least on the one-hand, and the difference between the most hierarchical image and the second-to-most, (2) the final ordered sets not to reverse the order of objective hierarchicalness of the images.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
51
array, and that, for each set, they would be required to select from among the images
displayed the one that they had seen in the “Images task”. Similar to Study 4, in order to
incentivize performance, we told (all) participants that the participant who had the most
correct responses would receive a substantial bonus ($7). Finally, because we wanted to
assess participants’ overall memory of the image rather than their ability to recall any
minor idiosyncratic differences across images, participants were told that the image array
would appear only for a short interval (45 seconds), and, thus, that they should rely on
their memory of the overall shape of the images rather than focusing on image details.
Participants then completed the memory task for each image set, with the sets
appearing in randomized order. For each of the four image sets, participants saw the focal
image, along with four distractor images: the two less hierarchical images and the two
more hierarchical ones (see Figure 3 for an example of one of the image sets). The
images within each set also appeared in randomized order, and all the images within each
set were presented using the same color background of the focal image in the “Images
task” (so as to mitigate against confusing images across sets). Participants had 45 seconds
to register their response, after which their response was coded as a missing value and the
next trial was automatically presented. After completing the memory task, participants
completed measures of their social conservatism and social dominance orientation (α =
.94), using the same items as in Study 3b.
Results and Discussion
We were centrally interested in whether participants’ levels of SDO could predict
the direction of errors they made in the memory task. Specifically, we wanted to know
whether individuals higher in SDO would misremember having seen less hierarchical
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
52
organizations than they actually had, and whether individuals lower in SDO would
misremember having seen more hierarchical organizations. Across the image sets, we
computed a score reflecting accuracy and direction of errors: specifically, for each of the
four image sets, participants were given a score of 0 if they correctly identified the image
they had previously seen, a score of -1 if they incorrectly indicated having seen a less
hierarchical image than they actually had, and a score of +1 if they incorrectly indicated
having seen a more hierarchical image. We summed participants’ scores across the four
image sets to compute our variable of hierarchy memory bias. On this index, negative
scores indicated an overall tendency towards underestimating the level of hierarchy one
had seen previously, and positive scores indicated a tendency towards overestimating the
same.16
On average, this was a difficult task for participants: 37.2% of participants did not
make the accurate selection for any of the four image sets, 45.2% made one accurate
selection, 14.8% made two accurate selections, and 2.8% made three accurate selections.
No participants were accurate for all four trials. On average, there was a slight bias
towards remembering less hierarchy than actually encountered (M = -.42, SD = 1.59, one-
sample t (572)= -6.38, p < .001). Neither of SDO (r = -.02, p = .62) nor social
16 Because we had a total of 5 organizations varying in hierarchy in each set, we also considered weighting each response by the level of hierarchy of the organization selected (i.e., -2 for incorrectly choosing the least hierarchical triangle, -1 for incorrectly choosing the second-least hierarchical organization, 0 for an accurate selection, +1 for incorrectly choosing the second-most hierarchical organization, and +2 for incorrectly choosing the most hierarchical organization). However, we chose not to pursue this as our primary analysis strategy because whereas our pretest ensured the selection of images with an equal difference in overall hierarchy between the set of more vs. less hierarchical organizations (relative to the focal organization), the difference in ratings of hierarchicalness between the most vs. second-most (and least vs. second-least) hierarchical organizations was not uniform (see also the previous footnote). Nevertheless, if we weight the responses in this way, we obtain similar results (see section 5 of the Supplemental Materials for full details).
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
53
conservatism (r = -.02, p = .60) were significantly associated with the number of accurate
trials.
We next turned to examine our central hypothesis. In zero-order terms, SDO was
negatively correlated with memory bias, r = -.11, p = .01, whereas social conservatism
was unassociated (r = .05, p = .24). When we entered SDO and social conservatism as
predictors into a simultaneous regression, we observed that SDO was negatively
associated with hierarchy memory bias (b = -.45, β = -.14, p = .003, 95% CI: [-.73, -.18]).
On the other hand, social conservatism was marginally positively associated with
memory bias (b = .08, β = .09, p = .07, 95% CI: [-.01, .17]), though this pattern is likely
best interpreted as a suppressor variable effect given that it social conservatism not
associated with hierarchy memory bias in zero-order terms. Thus, our results indicated
that the higher an individuals’ level of SDO, the greater their tendency to incorrectly
remember having seen less hierarchical organizations than they actually had seen (and the
less their tendency to misremember having seen more hierarchal organizations).
Importantly, our outcome measure simultaneously took into account trials in
which participants had overestimated hierarchy, had underestimated it, or had been
accurate. Thus, our patterns may have been driven by low SDO individuals
misremembering too much hierarchy, high SDO individuals misremembering too little, or
both. In order to distinguish between these possibilities, we created two additional
outcome measures: the first, labeled underestimation of hierarchy, simply indexed the
number of trials in which participants had selected a less hierarchical image than they had
actually seen. The second, labeled overestimation of hierarchy, indexed the number of
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
54
trials in which participants had selected a more hierarchical image than they had actually
seen.
We examined each of these outcomes separately. In zero-order terms, SDO was
significantly associated with underestimation of hierarchy (r = .11, p = .008) whereas
social conservatism was not (r = -.06, p = .15). When entered into a simultaneous Poisson
regression, SDO was significantly and positively associated with underestimation of
hierarchy (b = .19, Wald χ2 = 11.40, p = .001, 95% CI [.08, .29]). Although not correlated
in zero-order terms, social conservatism was associated with less underestimation (b = -
.04, Wald χ2 = 4.83, p = .03, 95% CI: [-.07, -.004]), indicating a likely suppressor
variable effect. In sum, the higher individuals were on SDO, the more likely they were to
misremember having seen less hierarchy than they had actually seen. On the other hand,
we observed that SDO was (marginally) negatively correlated with overestimating
hierarchy (r = -.07, p = .09), whereas social conservatism was unassociated (r = .02, p =
.57). When entered into a simultaneous Poisson regression, SDO was significantly
associated with overestimating hierarchy (b = -.16, Wald χ2 = 4.03, p = .045, 95% CI: [-
.31, -.004]), whereas social conservatism was unassociated (b = .02, Wald χ2 = 0.748, p =
.39, 95% CI [-.025, .066]). Thus, the lower individuals were on SDO, the more likely
they tended to be to overestimate hierarchy.17
In sum, we observed evidence consistent with our theorizing. On average, and
even when we financially incentivized accuracy, participants were prone to errors when
we asked them to remember which of a series of abstract depictions of hierarchical
organizations they had seen in a previous task. Importantly, individuals’ overall tendency
17 Because SDO was only marginally associated with overestimating hierarchy in zero-order terms, this relationship should be considered more tentative.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
55
to overestimate vs. underestimate the power differences in the organizations they had
seen could be predicted as a function of their orientation towards hierarchy-maintenance.
Specifically, individuals who prefer maintaining hierarchy in society remembered having
seen organizations that were less hierarchical than they really were. On the other hand,
there was a marginal tendency for those who reject hierarchy in society and prefer
egalitarian social arrangements to remember having seen more hierarchical organizations
than they actually had. Although it should be noted that these were small effects
(unsurprising, given the nature of the task and the degree of noise involved), the patterns
in the divergence between the memories of individuals higher vs. lower in SDO was
consistent across two separate large samples, here combined.18
Moreover, and importantly, this study allowed us to objectively assess perceptions
of hierarchy. Although we observed divergence in the perception of power differences as
a function of SDO in our prior studies, because these perceptions were inherently
subjective (i.e., there is no readily-discernible correct extent of power difference between
men and women, or Whites and Blacks), it was not possible to determine whether this
divergence was due to low SDO individuals exaggerating hierarchy, high SDO
individuals minimizing it, or both. Using our innovative biased memory methodology,
however, we were able to obtain evidence suggesting that both processes may be playing
a role. Finally, this study provided further evidence suggesting that the relationships we
observe likely do not reflect strategic misreporting of hierarchy, given that SDO was
18 We note that using the Stouffer method (Mostteller & Bush, 1954; see also Rosenthal and Rosnow (1991, p. 504)) to meta-analyze the effects from each of the two sub-studies combined here yielded the same conclusions. Specifically, SDO was significantly correlated with the hierarchy memory bias (weighted z = 2.51, p = .01) and underestimating hierarchy (weighted z = 2.59, p = .009), and marginally correlated with overestimating hierarchy (weighted z = 1.71, p = .087).
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
56
shown to relate to a measure of one’s memory of hierarchy (as opposed to just ratings of
it), and with their accuracy again financially incentivized.
General Discussion
Consensus on how to respond to inequality will remain elusive without agreement
on how much of it there is in the first place. Yet the research presented here implies that
people come to perceive systematically different levels of inequality depending on how
desirable they believe it to be. Across eight studies (and four supplementary studies), we
found evidence highly consistent with the notion that individuals’ motivational
orientation towards hierarchy influences their perception of the degree of social
inequality: Individuals motivated to maintain hierarchy (i.e., those high in social
dominance orientation or other measures of anti-egalitarianism) perceive less inequality,
whereas those motivated to attenuate hierarchy perceive more.
This association between equality-motives and hierarchy perception applies
broadly: We observed the above pattern among members of both advantaged and
disadvantaged groups, across a wide range of contexts (both real-world and fictional),
and using a variety of stimuli, methods, and measures of egalitarianism (see Study 3c).
Anti-egalitarianism was associated —among both men and women, blacks and whites,
and rich and poor— with perceiving less hierarchy between groups in real-world settings
(Studies 1a, 1b, and Supplementary Study 1). The relationship between SDO and
hierarchy-perceptions was also observed when we asked participants to read about an
asymmetric intergroup conflict between novel, fictitious groups (Study 2 and
Supplementary Study 2). We obtained strikingly similar results when we depicted power
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
57
relations using abstract visual diagrams, whether we examined perception (Studies 3a-4
and Supplementary Studies 3-4) or recognition memory (Study 5).
Beyond increasing our confidence in the robustness of the association between
egalitarianism and inequality perceptions, our examination of this relationship across a
range of methods and modalities allowed us to arbitrate between several theoretical
accounts that could have explained our findings.
One possible explanation for the relationship between SDO and perception of
hierarchy is that SDO is also associated with the kind of opportunities one might have to
learn about group-based inequality. From this perspective, the results obtained in Studies
1a and 1b may be due to those high and low in SDO encountering different underlying
information about hierarchy in the world, for example via self-selection into jobs or
neighborhoods where power differences between groups happened to be less vs. more
visible, respectively. Thus, it is possible that individuals high and low in SDO end up
differing in their perception of inequality between racial, gender, and class groups only
because their environments differ in the information they provide about group-based
disparities (and not because of any biased information processing on individuals’ part).
Studies 2-5 allowed us to rule out this possibility, as they involved presenting all
participants with exactly the same information about hierarchy: though everyone read the
same text about an asymmetric intergroup context in Study 2 (and Supplementary Study
2), and saw the same abstract images of hierarchical societies and organizations in
Studies 3a-5 (and Supplementary Studies 3-4), the association between SDO and
perception of hierarchy persisted. This occurred despite the fact that these studies focused
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
58
on novel groups and organizations with which participants could not have had prior
context-specific experiences that might have shaped their perceptions.
The second possibility we considered was that any differences in ratings of
hierarchy reflected not intuitive perceptual biases, but conscious strategic minimization
or maximization of hierarchy, intended to legitimize individuals’ favored policy leanings.
If participants were simply electing to strategically misreport the levels of hierarchy they
were perceiving, then providing a strong countervailing incentive might have been
expected to convince them to report their true perceptions. However, randomly
incentivizing one group of participants (but not the other) to accurately report their power
perceptions in Study 4 (and Supplementary Study 4) had no effect on the relationship
between SDO and power perceptions, despite the fact that a pretest found this incentive
to increase the desire to respond accurately. Combined with the fact low and high SDO
participants (again incentivized to be accurate) in Study 5 incorrectly remembered having
seen more or less hierarchal organizations, this suggests that participants may have truly
biased perceptions of inequality, seeing what they want to see as a function of their
underlying motivation toward vs. against hierarchy.
More than differential exposure to underlying information or strategic reporting,
then, our results support the possibility that individuals are in fact processing information
about hierarchy to which they were exposed in a biased manner. This is in line with
research on motivated information processing (Balcetis & Dunning, 2013; Chambers et
al., 2014, 2015; Crawford et al., 2013; Ditto et al., 2009; Munro & Ditto, 1997; Hennes et
al., under review; 2015-b; Kahan, 2012; Kteily et al., 2014; Taber & Lodge, 2006), to
which our findings make several contributions.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
59
In assessing perceptions of hierarchy using pictorial depictions of societies and
organizations, our work is among the few studies to examine how motives related to
complex real-world political constructs (i.e., social hierarchy) are associated with the
processing of even highly abstracted (though still motivationally-relevant) stimuli (see
also Okimoto & Gromet, in press). Our work thus provides evidence supporting the idea
that individuals’ political motives may be quite powerful, potentially influencing the
processing of even very basic stimuli that lack the richness of stimuli previously
considered in this context (e.g., newspaper articles favoring one political view over
another, or videos of protest behavior; Crawford et al., 2013; Kahan et al., 2012):
Individuals’ motives may lead them not only to dismiss the validity of news content that
disagrees with their views, but even to rate abstract organizational hierarchies as looking
‘flatter’ or ‘steeper’.
The memory paradigm we introduced in Study 5 also makes important
contributions to research on motivated perception, on both the conceptual and empirical
levels. First, by showing that our findings extended to the domain of memory, we
contribute to the limited existing research on motivated memory biases, adding equality-
motives to a short list of social-psychological motivations known to influence the
information individuals remember (see also Hennes et al, under review; Story, 1988;
Sanitioso, Kunda, & Fong, 1990). Second, the empirical paradigm itself has advantages
that may be of use to other researchers. By asking participants to correctly recall which of
a series of organizations (that systematically varied in their hierarchy) they had
previously seen, we were able to create an objective metric against which to assess
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
60
accuracy, providing clarity about whether our effects were driven by individuals high (vs.
low) in our independent variable (or both).
Finally, our work extends research suggesting that motivated perceptions can
have important implications (Riccio, Cole, & Balcetis, 2013). Indeed, across several
studies in real-world and fictional contexts, we consistently found that perceptions of
power differences mediated the effects of SDO on support for social policy. Our findings
are consistent with the idea that individuals whose low (high) SDO leads them to actually
see more (less) hierarchy subsequently feel a greater (lesser) need to enact policies
intended to equalize group-based differences. Whereas there is a good deal of research
suggesting that individuals’ motivations can lead them to perceive the world differently,
there is less research documenting downstream implications of these perceptions (see
also Okimoto & Gromet, in press).
Beyond its contributions to research on motivated cognition, our work also has
several implications for theorizing in intergroup relations, and social dominance theory in
particular. One interesting observation was the fact that levels of SDO predicted
perceptions of inequality between each of racial, gender, and class groups (which were
highly inter-correlated and formed a unidimensional scale). Whereas prior work on
motivated biases in perceptions of hierarchy focused on one specific context, such as race
(e.g., Kahn, Ho, Sidanius, & Pratto, 2009) or class (e.g., Chambers et al., 2014, 2015;
Kraus & Tan, 2015), we reasoned that, as an orientation towards hierarchy broadly
considered, SDO would predict biased perceptions in a wide range of hierarchy-relevant
contexts. Thus, our findings are consistent with the conceptualization of SDO as a
general orientation towards hierarchy (Ho et al., 2015; Kteily et al., 2012; Sibley & Liu,
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
61
2010), rather than simply reflecting feelings about (or membership within) any particular
group categorization (Schmitt, Branscombe, & Kappen, 2003; see also Kteily, Sidanius,
& Levin, 2011).
Our finding that the link between SDO and perceptions of hierarchy tended not to
differ as a function of membership in advantaged (e.g., Whites, men, high SES
individuals) vs. disadvantaged (e.g., Blacks, women, low SES individuals) groups is also
theoretically noteworthy.19 First, it supports theoretical perspectives arguing that
individuals’ views about the social system are shaped not only by the social standing of
the groups to which they belong (Tajfel & Turner, 1979), but also by their ideological
motives (Sidanius & Pratto, 1999; Jost et al., 2004). Thus, high SDO subordinate group
members under-perceive hierarchy despite the fact that it is not in their group interest to
do so, consistent with the argument in system justification theory that disadvantaged
group members are sometimes motivated to defend the societal status quo that
disadvantages them (Jost, Banaji, & Nosek 2004; see also Sidanius & Pratto, 1999, for a
discussion of ‘behavioral asymmetry’, which describes how members of low status
groups sometimes hold beliefs or behave in ways that harm the ingroup). Second, this
finding is also consistent with the view that SDO reflects generalized support for
hierarchy rather than simply reflecting desire for the dominance of the ingroup (i.e., Ho et
al., 2015; Pratto & Stewart, 2012).
19 Although we focused on examining the moderating effect of ethnicity, class, and gender in Studies 1a, 1b and Supplementary Study 1, where we expected them to be most relevant (because of these studies’ emphasis on hierarchy perceptions within these three contexts), we also examined the potential moderating effect of these variables in the remaining studies (see section 4 of the Supplemental Materials for full details). Consistent with the conclusions in Studies 1a, 1b, and Supplementary Study 1, we observed little evidence of reliable moderation in these studies (across the 30 tests we conducted, only 4 were significant).
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
62
Furthermore, the finding that perceptions of hierarchy mediate effects of SDO on
policy support importantly extends social dominance theory’s concept of hierarchy-
enhancing and attenuating beliefs (Sidanius & Pratto, 1999), to encompass not only
ideologies but also perceptual biases. Previous work suggests that high and low SDO
individuals reject or support egalitarian social policies in part because they endorse
hierarchy-enhancing (e.g., Protestant worth ethic; racism) or attenuating (e.g., socialism;
noblesse oblige) ideologies, which rationalize or problematize existing levels of
hierarchy. The present work suggests another reason: high and low SDO individuals may
also come to display biased perceptions of the degree of hierarchy, which subsequently
reinforce their convictions about the types of social policies they tend to favor. The
notion of hierarchy-enhancing and attenuating perceptual biases is consistent with recent
evidence that individuals’ SDO motives influence their perception of racially ambiguous
targets in ways that help maintain or attenuate hierarchy (see Ho et al., 2013; Kteily et al.,
2014; McClanahan, Kteily, & Ho, under review).
Such perceptual biases could be particularly resistant to persuasion, entrenching
polarization and increasing frustration in political debates: whereas it may be possible (if
challenging) to present compelling logical arguments that lead individuals to reconsider
their support for a particular policy position, it may be more difficult for two individuals
who perceive the same reality differently to establish common ground. This is likely to be
especially difficult if individuals are unaware that their perception is biased. It is
noteworthy in this regard that participants continued to perceive hierarchy differently
even when we experimentally induced a desire for accuracy in Study 4. This is consistent
with research on the involuntary nature of cognitive biases, which regularly occur outside
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
63
conscious awareness and may thus be insensitive to financial incentives (Kahneman,
2003; see also Camerer & Hogarth, 1999).20
Finally, because the memory paradigm we developed allowed us to investigate
whether the correlation between SDO and hierarchy perception was a sign of distorted
perception of those high vs. low in SDO, we were able to contribute to the debate on
whether motivated cognition is a feature of those with right-wing attitudes, left-wing
attitudes, or both (see also Bakshy et al., 2015; Barberá et al., 2015; Chambers et al.,
2014, 2015; Kahan et al., 2012). We found qualified support for the bidirectional
perspective: there was a significant relationship showing that those higher in SDO
remembered less hierarchy than objectively experienced, whereas we observed a
marginally significant relationship such that those lower in SDO remembered more.21
Given the relationship between SDO and economic conservatism (Jost et al., 2009), our
results suggest that any attempts to addressed biased perception of inequality in society
will likely need to target those at both ends of the ideological spectrum.
Limitations and Future Directions
Notwithstanding the contributions made by the present research, several questions
are worth considering further. For one, it should be acknowledged that a majority of our
studies (though not all; see Study 1a and Supplementary Study 1) employed samples
recruited from Amazon’s mTurk platform. Although mTurk is known to be a reliable and
20 Interestingly, this work suggests that one method which can sometimes be effective in attenuating unconscious cognitive and perceptual biases is to make individuals aware of how they work (Kahneman, 2003). However, the biases this research refers to (e.g., the accessibility heuristic; ignorance of base-rates) tend not to have the ideological aspects of the bias we consider here. Thus, it remains to be seen whether informing individuals of the association between their political motives and their hierarchy perception would be effective in reducing biased perception. 21 We examined this question among a relatively large sample (n = 577). Nevertheless, because the effects in our memory paradigm were relatively small, these conclusions should be considered tentative until future work has replicated the effects with still larger samples.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
64
high-quality platform for recruiting relatively diverse participant samples (Buhrmester,
Kwang, & Gosling, 2011), participants on this platform are known to be relatively
experienced study-takers (Rand et al., 2014) and to differ in certain respects from the
general population (e.g., exhibiting greater anxiety; Arditte, Çek, Shaw, & Timpano, in
press). Thus, it would be worthwhile for future work to examine the patterns we observe
in other populations.
From a theoretical perspective, future work would benefit from more directly
isolating the role of motivation in the perceptual bias we identify. Consistent with a large
body of work on motivated cognition and perception examining the influence of chronic
motivations (e.g., Chambers et al., 2014, 2015; Cole, Balcetis & Dunning, 2013;
Crawford et al., 2013; Dijksterhuis, Van Knippenberg, Kruglanski, & Schaper, 1996;
Epley & Whitchurch, 2008; Gollwitzer, Kappes, & Oettingen, 2011; Isaacowitz,
Wadlinger, Goren, & Wilson, 2006; Makhanova, Miller, & Maner, 2015), we measured
individual differences in equality motives. Our studies also ruled out two alternative
explanations for the relationship we observed that do not rely on motivated perceptual
biases: encountering different underlying information about hierarchy, and strategically
misreporting perceptions. Nevertheless, because we do not directly manipulate
motivation in our studies, we cannot definitively rule out all alternative explanations
(although we are not aware of alternative non-motivational explanations that could
readily account for the patterns we observe).
Beyond manipulating individuals’ preference for equality between groups, future
work could provide further support for our motivated account by examining the SDO-
equality perception link cross-culturally. We reasoned that egalitarian social norms in the
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
65
U.S. (Kahn et al., 2009; Katz & Hass, 1988) create social pressures for attenuating
hierarchy when it is extensive, generating an incentive among high SDO individuals to
under-perceive hierarchy (i.e., making them want to see the ‘flatness’ that justifies
avoiding egalitarian change) and low SDO individuals to over-perceive it (i.e., making
them want to see the ‘steepness’ that justifies corrective action). If this reasoning is
correct, one way to isolate the role of motivation would be to compare our effects in the
U.S. with those in other cultural contexts without such norms. In contexts where
hierarchy is perceived as natural or even desirable (e.g., India, China; Hofstede, Hofstede,
& Minkov, 1997), high levels of inequality might not generate pressure for egalitarian
policy. In such settings, our theorizing would predict a reduced incentive among high
SDO individuals to under-perceive hierarchy and among low SDO individuals to over-
perceive it. Such evidence would provide strong support for our motivated account.
Although we did not have such cross-cultural evidence in the present work, we
did conduct supplementary analyses that bolster our motivated account. We reasoned that
if individuals’ biased perception is stemming from their motivation to over-perceive (low
SDO) and under-perceive (high SDO) hierarchy in response to social pressures, we
should observe a fairly specific pattern of bias. Specifically, high SDO individuals’
motive to under-perceive hierarchy should be stronger the greater hierarchy is (and thus,
the stronger social pressure to reduce it might be). Conversely, low SDO individuals
should show a stronger motivation to over-perceive hierarchy the less hierarchy there is
(and thus, the more likely others in society will deem it unproblematic and ignore it).
Following this logic, one would expect the perceptions of low and high SDO individuals
to deviate systematically from those with more moderate levels, with high SDO
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
66
individuals’ perceptions diverging more from those of the average person the more
hierarchical society becomes, and low SDO individuals showing the opposite pattern (see
Figure 4). We examined this possibility with the data from Studies 3b and 4, in which we
had a large group of participants rate the hierarchy of five organizations that varied in
their degree of hierarchy. We find that the pattern of divergence in the ratings among low
and high (vs. average) SDO individuals tracks the theoretical pattern our reasoning would
predict (see section 6 of the Supplemental Materials for full details).
One important job for future research is to further specify the mechanisms by
which individuals’ motivations result in their biased perception of hierarchy. For
example, future work could explore exactly how individuals’ equality motives influenced
their perceptions of the organizations depicted in Studies 3b-5. One possibility is that
individuals directed their attention to different parts of the image. Thus, high SDO
individuals might have focused on the large number of individuals at the bottom of the
organizational pyramid, interpreting them as a formidable potential threat to those at the
top; low SDO individuals might instead have focused their attention on the large number
of layers that those at the bottom would need to traverse to get to the top (see Riccio et
al., 2013, for a discussion of directed attention). It would be interesting to examine this
question using eye-tracking technology, to see whether individuals do indeed fixate on
different parts of the image in ways that might contribute to the perceptual impression
they form (see Balcetis & Dunning, 2006 for a similar approach).
Processing information from different vantage points might similarly have
contributed to the findings in Study 2 (and Supplemental Study 2), even if visual
perception was not involved. For example, in reading about the dispute between the
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
67
(advantaged) Group A and (disadvantaged) Group B in Study 2, high SDO individuals
might have anchored on the challenges involved in maintaining power in the face of
resistance (leading them to see less of an advantage for Group A), whereas low SDO
individuals might have anchored on the relative deprivation facing the members of Group
B, who were said to have lost access to the majority of resources in ‘Raga’ (making them
see Group B as more disadvantaged). Future studies that assess (or manipulate) the
interpretive frames that individuals apply in processing information about hierarchy could
shed light on this possibility (see also Eibach & Ehrlinger, 2006). Lastly, it would be
interesting to employ longitudinal methods to examine whether, over time, the news
articles or television programs that individuals high vs. low in SDO choose to engage
with contribute to differences in their perceptions of group-based disparities in the real-
world contexts we examined in Studies 1a and 1b (and Supplementary Study 1).
Future work could also more extensively test the causal order of the variables in
our model. Our theorizing proposes that individuals’ social dominance motives influence
their perceptions of hierarchy between groups, rather than the reverse. This is consistent
with the widespread view of SDO as a stable individual difference variable that predicts
downstream attitudes and behavior and that, once formed, is relatively (though not
completely) impervious to change (e.g., Duckitt, 2001; Kteily et al., 2011; see Dhont,
Van Hiel, & Hewstone, 2014). In Study 1b, we conducted a longitudinal study examining
whether levels of SDO predicted perceptions of hierarchy between real-world groups 4-6
weeks later, controlling for original levels of perceived hierarchy. In line with our
perspective, the results of this study provided evidence for an effect of SDO on hierarchy
perceptions over time; on the other hand, perceptions of hierarchy did not have a
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
68
longitudinal effect on SDO. At the same time, because this study examined a short period
of time, its results are best interpreted as suggestive rather than definitive. Future work
should employ longitudinal designs over a longer time course, and could further consider
examining the cross-lagged relationship between SDO and perceptions of hierarchy
during ‘impressionable years’, or periods of time (e.g., adolescence; early adulthood) in
which individuals’ sociopolitical beliefs are thought to be more malleable (Alwin et al.,
1991). Relatedly, future research could examine the effects on hierarchy perceptions of
experimentally manipulating levels of SDO (see Guimond, Dambrun, Michinov &
Duarte, 2003).
Finally, it would be good for future work to simultaneously examine the roles of
perceptual biases and legitimizing ideologies in influencing policy attitudes. We argue
here that, in addition to adopting ideologies that legitimize their stance on whether
existing levels of hierarchy are problematic, individuals may perceive the degree of
hierarchy itself in ways that justify their policy stances. Future studies could examine
how each of perceiving hierarchy and explaining it are inter-related, and whether each
uniquely mediates the effects of SDO on egalitarian policy support.
Conclusion
Across a range of studies, we provide evidence consistent with the idea that
individuals’ support vs. opposition to hierarchy biases the basic ways in which they
perceive the extent of inequality between social groups. By potentially changing the very
way in which we perceive hierarchy-relevant information, individual differences in our
orientation toward group-based hierarchy may even come to influence attitudes toward
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
69
egalitarian social policies, thus contributing to the intensity and intractability of debates
about how to structure our societies.
References
Arditte, K. A., Çek, D., Shaw, A. M., & Timpano K. R. (in press). The Importance of
Assessing Clinical Phenomena in Mechanical Turk Research. Psychological
Assessment.
Baker, P., & Fausset, R. (2015, March 7). Obama, at Selma memorial, says, “We know
the march is not yet over”. New York Times. Retrieved from
http://www.nytimes.com/2015/03/08/us/obama-in-selma-for-edmund-pettus-
bridge-attack-anniversary.html
Bakshy, E., Messing, S., & Adamic, L. (2015). Exposure to ideologically diverse news
and opinion on Facebook. Science, 348, 1130-1132.
Balcetis, E., & Dunning, D. (2006). See what you want to see: motivational influences on
visual perception. Journal of Personality and Social Psychology, 91(4), 612-625.
Balcetis, E., & Dunning, D. (2007). Cognitive dissonance and the perception of natural
environments. Psychological Science, 18(10), 917-921.
Balcetis, E. & Dunning, D. (2013). Considering the situation: Why people are better
social psychologists than self-psychologists. Self and Identity, 12(1), 1-15.
Barak, M. E. M. (2013). Managing diversity: Toward a globally inclusive workplace.
Thousand Oaks, CA: Sage.
Barberá, P., Jost, J. T., Nagler, J., Tucker, J. A., & Bonneau, R. (2015). Tweeting From
Left to Right Is Online Political Communication More Than an Echo
Chamber? Psychological Science, 26 (10), 1531-1542 .
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
70
Blow, C. M. (2013, September 13). Occupy Wall Street Legacy. New York Times.
Retrieved from http://www.nytimes.com/2013/09/14/opinion/blow-occupy-wall-
street-legacy.html.
Bobo, L., & Kluegel, J. R. (1993). Opposition to race-targeting: Self-interest,
stratification ideology, or racial attitudes? American Sociological Review, 58(4),
443-464.
Bobocel, D. R., Son Hing, L. S., Davey, L. M., Stanley, D. J., & Zanna, M. P. (1998).
Justice-based opposition to social policies: Is it genuine? Journal of Personality
and Social Psychology, 75(3), 653-669.
Brandt, M. J., Reyna, C., Chambers, J. R., Crawford, J. T., & Wetherell, G. (2014). The
ideological-conflict hypothesis intolerance among both liberals and conservatives.
Current Directions in Psychological Science, 23(1), 27-34.
Brodish, A. B., Brazy, P. C., & Devine, P. G. (2008). More eyes on the prize: Variability
in White Americans’ perceptions of progress toward racial equality. Personality
and Social Psychology Bulletin, 34, 513-527.
Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazon's Mechanical Turk a new
source of inexpensive, yet high-quality, data?. Perspectives on psychological
science, 6(1), 3-5.
Camerer, C. F., & Hogarth, R. M. (1999). The effects of financial incentives in
experiments: A review and capital-labor-production framework. Journal of risk
and uncertainty, 19(1-3), 7-42.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
71
Chambers, J. R., Swan, L. K., & Heesacker, M. (2014). Better off than we know:
Distorted perceptions of incomes and income inequality in America.
Psychological Science, 25(2), 613-618.
Chambers, J. R., Swan, L. K., & Heesacker, M. (2015). Perceptions of U.S. social
mobility are divided (and distorted) along ideological lines. Psychological
Science, 26, 413-423.
Chetty, R., Hendren, N., Kline, P., & Saez, E., & Turner, N. (2014). Is the United States
Still a Land of Opportunity? Recent Trends in Intergenerational Mobility. NBER
Working Paper No. 19844. Retrieved from http://www.equality-of-
opportunity.org/files/mobility_trends.pdf on 7/28/15.
Cohn, N. (2014, June 12). Polarization is dividing American society, not just politics.
New York Times. Retrieved from
http://www.nytimes.com/2014/06/12/upshot/polarization-is-dividing-american-
society-not-just-politics.html?abt=0002&abg=0
Cole, S., Balcetis, E., & Dunning, D. (2013). Affective signals of threat increase
perceived proximity. Psychological Science, 24(1), 34-40.
Cotterill, S., Sidanius, J., Bhardwaj, A., & Kumar, V. (2014). Ideological support for the
Indian caste system: Social dominance orientation, right-wing authoritarianism
and karma. Journal of Social and Political Psychology, 2(1), 98-116.
Crawford, J. T., Jussim, L., Cain, T. R., & Cohen, F. (2013). Right‐ wing
authoritarianism and social dominance orientation differentially predict biased
evaluations of media reports. Journal of Applied Social Psychology, 43(1), 163-
174.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
72
Cronbach, L. J., & Furby, L. (1970). How we should measure" change": Or should
we?. Psychological bulletin, 74(1), 68-80.
Dawtry, R. J., Sutton, R. M., & Sibley, C. G. (in press). Why wealthier people think
people are wealthier, and why it matters: From social sampling to redistributive
attitudes. Psychological Science.
Dijksterhuis, A. P., Van Knippenberg, A. D., Kruglanski, A. W., & Schaper, C. (1996).
Motivated social cognition: Need for closure effects on memory and judgment.
Journal of Experimental Social Psychology, 32(3), 254-270.
Ditto, P. H., & Lopez, D. F. (1992). Motivated skepticism: use of differential decision
criteria for preferred and nonpreferred conclusions. Journal of Personality and
Social Psychology, 63(4), 568-584.
Ditto, P. H., Pizarro, D. A., & Tannenbaum, D. (2009). Motivated moral
reasoning. Psychology of Learning and Motivation, 50, 307-338.
Dhont, K., Hodson, G., Costello, K., & MacInnis, C. C. (2014). Social dominance
orientation connects prejudicial human–human and human–animal
relations. Personality and Individual Differences, 61, 105-108.
Dhont, K., Van Hiel, A., & Hewstone, M. (2014). Changing the ideological roots of
prejudice: Longitudinal effects of ethnic intergroup contact on social dominance
orientation. Group Processes & Intergroup Relations, 17 (1), 27-44.
Dubois, D., Rucker, D. D., & Galinsky, A. D. (2015). Social class, power, and
selfishness: When and why upper and lower class individuals behave
unethically. Journal of Personality and Social Psychology, 108(3), 436-449.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
73
Duckitt, J. (2001). A dual-process cognitive-motivational theory of ideology and
prejudice. Advances in Experimental Social Psychology, 33, 41–113.
Duckitt, J., & Sibley, C. G. (2010). Personality, ideology, prejudice, and politics: A dual
process motivational model. Journal of Personality, 78(6), 1861-1894.
Edwards, J. R. (1994). Regression analysis as an alternative to difference scores. Journal
of Management, 20(3), 683-689.
Eibach, R. P., & Ehrlinger, J. (2006). “Keep your eyes on the prize”: Reference points
and racial differences in assessing progress toward equality. Personality and
Social Psychology Bulletin, 32(1), 66-77.
Eibach, R. P., & Keegan, T. (2006). Free at last? Social dominance, loss aversion, and
White and Black Americans' differing assessments of racial progress. Journal of
Personality and Social Psychology, 90(3), 453-467.
Emerson, R. M. (1962). Power-dependence relations. American Sociological Review, 31-
41.
Epley, N., & Whitchurch, E. (2008). Mirror, mirror on the wall: Enhancement in self-
recognition. Personality and Social Psychology Bulletin, 34(9), 1159-1170.
Eriksson, K., & Simpson, B. (2012). What do Americans know about inequality? It
depends on how you ask them. Judgment and Decision Making, 7, 741-745.
Frey, D. (1986). Recent research on selective exposure to information. Advances in
Experimental Social Psychology, 19(1), 41-80.
Garrett, R. K. (2009). Echo chambers online?: Politically motivated selective exposure
among Internet news users. Journal of Computer‐ Mediated
Communication, 14(2), 265-285.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
74
Goodman, E., Adler, N. E., Kawachi, I., Frazier, A. L., Huang, B., & Colditz, G. A.
(2001). Adolescents’ perceptions of social status: Development and evaluation of
a new indicator. Pediatrics, 108, 1–8.
Gollwitzer, P. M., Kappes, H. B., & Oettingen, G. (2011). Needs and incentives as
sources of goals. Goal-directed behavior, 115-149.
Grant, A., & Sandberg, S. (2014, Dec 6). When talking about bias backfires. New York
Times. Retrieved from
http://www.nytimes.com/2014/12/07/opinion/sunday/adam-grant-and-sheryl-
sandberg-on-discrimination-at-work.html
Guimond, S., Dambrun, M., Michinov, N., & Duarte, S. (2003). Does social dominance
generate prejudice? Integrating individual and contextual determinants of
intergroup cognitions. Journal of Personality and Social Psychology, 84(4), 697-
721.
Guimond, S., De Oliveira, P., Kamiesjki, R., & Sidanius, J. (2010). The trouble with
assimilation: Social dominance and the emergence of hostility against
immigrants. International Journal of Intercultural Relations, 34(6), 642-650.
Guardian (2015, June 10). Nobel scientist Tim Hunt: Female scientists cause trouble for
me in labs. The Guardian. Retrieved from http://www.theguardian.com/uk-
news/2015/jun/10/nobel-scientist-tim-hunt-female-scientists-cause-trouble-for-
men-in-labs
Hart, W., Albarracín, D., Eagly, A. H., Brechan, I., Lindberg, M. J., & Merrill, L. (2009).
Feeling validated versus being correct: a meta-analysis of selective exposure to
information. Psychological Bulletin, 135(4), 555-588.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
75
Hatemi, P. K., Medland, S. E., Klemmensen, R., Oskarsson, S., Littvay, L., Dawes, C. T.,
... & Christensen, K. (2014). Genetic influences on political ideologies: Twin
analyses of 19 measures of political ideologies from five democracies and
genome-wide findings from three populations. Behavior Genetics, 44(3), 282-294.
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process
analysis: A regression-based approach. Guilford Press.
Ho, A. K., Sidanius, J., Pratto, F., Levin, S., Thomsen, L., Kteily, N., & Sheehy-
Skeffington, J. (2012). Social dominance orientation revisiting the structure and
function of a variable predicting social and political attitudes. Personality and
Social Psychology Bulletin, 38(5), 583-606.
Ho, A. K., Sidanius, J., Cuddy, A. J. C., & Banaji, M. R. (2013). Status-boundary
enforcement and the categorization of Black-White biracials. Journal of
Experimental Social Psychology, 49(5), 940-943.
Ho, A. K., Sidanius, J., Kteily, N., Sheehy-Skeffington, J, Pratto, F., Henkel, K. E., Foels,
R., & Stewart, A. L. (2015). The nature of social dominance orientation:
Theorizing and measuring preferences for intergroup inequality using the new
SDO7 scale. Journal of Personality and Social Psychology, 109(6), 1003-1028.
Hofstede, G., Hofstede, G. J., & Minkov, M. (1997). Cultures and organizations. New
York, NY: McGraw Hill.
Isaacowitz, D. M., Wadlinger, H. A., Goren, D., & Wilson, H. R. (2006). Selective
preference in visual fixation away from negative images in old age? An eye-
tracking study. Psychology and aging, 21(1), 40-48.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
76
Iyengar, S., & Hahn, K. S. (2009). Red media, blue media: Evidence of ideological
selectivity in media use. Journal of Communication, 59(1), 19-39.
Jost, J. T., Banaji, M. R., & Nosek, B. A. (2004). A decade of system justification theory:
Accumulated evidence of conscious and unconscious bolstering of the status
quo. Political Psychology, 881-919.
Jost, J. T., Federico, C. M., & Napier, J. L. (2009). Political ideology: Its structure,
functions, and elective affinities. Annual review of psychology, 60, 307-337.
Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F. J. (2003). Political conservatism
as motivated social cognition. Psychological Bulletin, 129(3), 339-375.
Judd, C. M., Westfall, J., & Kenny, D. A. (2012). Treating stimuli as a random factor in
social psychology: a new and comprehensive solution to a pervasive but largely I
gnored problem. Journal of Personality and Social Psychology, 103(1), 54-69.
Kahan, D. M., Hoffman, D. A., Braman, D., & Evans, D. (2012). They saw a protest:
Cognitive illiberalism and the speech-conduct distinction. Stan. L. Rev., 64, 851.
Kahan, D. M. (2012). Ideology, motivated reasoning, and cognitive reflection: an
experimental study. Judgment and Decision Making, 8, 407-24.
Kahn, K., Ho, A. K., Sidanius, J., & Pratto, F. (2009). The space between us and them:
Perceptions of status differences. Group Processes & Intergroup Relations, 12(5),
591-604.
Kahneman, D. (2003). A perspective on judgment and choice: mapping bounded
rationality. American psychologist, 58(9), 697-720.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
77
Kandler, C., Bleidorn, W., & Riemann, R. (2012). Left or right? Sources of political
orientation: The roles of genetic factors, cultural transmission, assortative mating,
and personality. Journal of Personality and Social Psychology, 102(3), 633-645.
Katz, I., & Hass, R. G. (1988). Racial ambivalence and American value conflict:
Correlational and priming studies of dual cognitive structures. Journal of
Personality and Social Psychology, 55(6), 893-905.
Keltner, D., Gruenfeld, D. H., & Anderson, C. (2003). Power, approach, and
inhibition. Psychological Review, 110(2), 265-284.
Knight, K. (1999). Liberalism and conservatism. In J. P. Robinson, P. R. Shaver, & L. S.
Wrightsman (Eds.), Measures of political attitudes (pp. 59 –158). San Diego, CA:
Academic Press.
Knowles, E. D., & Lowery, B. S. (2012). Meritocracy, self-concerns, and Whites' denial
of racial inequity. Self and Identity, 11(2), 202-222.
Kraus, M. W., & Tan, J. J. (2015). Americans overestimate social class mobility. Journal
of Experimental Social Psychology, 58, 101-111.
Kteily, N. S., Ho, A. K., & Sidanius, J. (2012). Hierarchy in the mind: The predictive
power of social dominance orientation across social contexts and domains.
Journal of Experimental Social Psychology, 48(2), 543-549.
Kteily, N. S., Cotterill, S., Sidanius, J., Sheehy-Skeffington, J., & Bergh, R. (2014). “Not
one of us”: Predictors and consequences of denying ingroup characteristics to
ambiguous targets. Personality and Social Psychology Bulletin, 40, 1231-1247.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
78
Kteily, N. S., Saguy, T., Sidanius, J., & Taylor, D. M. (2013). Negotiating power:
Agenda ordering and the willingness to negotiate in asymmetric intergroup
conflicts. Journal of Personality and Social Psychology, 105(6), 978-995.
Kteily, N. S., Sidanius, J., & Levin, S. (2011). Social dominance orientation: Cause or
‘mere effect’? Evidence for SDO as a causal predictor of prejudice and
discrimination against ethnic and racial outgroups. Journal of Experimental Social
Psychology, 47, 208-214.
Lord, C. G., Ross, L., & Lepper, M. R. (1979). Biased assimilation and attitude
polarization: The effects of prior theories on subsequently considered evidence.
Journal of Personality and Social Psychology, 37(11), 2098-2109.
Lowery, B. S., Knowles, E. D., & Unzueta, M. M. (2007). Framing inequity safely:
Whites’ motivated perceptions of racial privilege. Personality and Social
Psychology Bulletin, 33, 1237-1250.
Makhanova, A., Miller, S. L., & Maner, J. K. (2015). Germs and the out-group: Chronic
and situational disease concerns affect intergroup categorization. Evolutionary
Behavioral Sciences, 9(1), 8-19.
McClanahan, K. J., Kteily, N. S., & Ho, A. K. Cutting off Reflected Glory: Egalitarian
Whites Credit the Outgroup Over the Ingroup for High Status Ambiguous
Targets. Manuscript submitted for publication.
McCloskey, H., & Bann, C. A. (1979). On the reappraisal of the classical conservatism
scale. Political Methodology, 6, 149-172.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
79
Munro, G. D., & Ditto, P. H. (1997). Biased assimilation, attitude polarization, and affect
in reactions to stereotype-relevant scientific information. Personality and Social
Psychology Bulletin, 23(6), 636-653.
Murrell, A. J., Dietz-Uhler, B. L., Dovidio, J. F., Gaertner, S. L., & Drout, C. (1994).
Aversive racism and resistance to Affirmative Action: Perception of justice are
not necessarily color blind. Basic and Applied Social Psychology,15, 71-86.
New York Times (2011, November 9). Public opinion and the Occupy Movement. New
York Times. Retrieved from
http://www.nytimes.com/interactive/2011/11/09/us/ows-grid.html?_r=0
New York Times (2013, November 20). Racism in the age of Obama. New York Times.
Retrieved from http://www.nytimes.com/roomfordebate/2013/11/20/racism-in-
the-age-of-obama
Norton, M. I., & Ariely, D. (2011). Building a better America—One wealth quintile at a
time. Perspectives on Psychological Science, 6(1), 9-12.
Okimoto, T. G., Gromet, D. M. (in press). Differences in sensitivity to deviance partly
explain ideological divides in social policy support. Journal of Personality and
Social Psychology.
Porter, E. (2014, May 13). The politics of income inequality. New York Times. Retrieved
from http://www.nytimes.com/2014/05/14/business/economy/the-politics-of-
incomeinequality.html?gwh=ADE07668FDFC30F4981F8607DF3B675F&gwt=p
ay&_r=0
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
80
Pratto, F., Sidanius, J., & Levin, S. (2006). Social dominance theory and the dynamics of
intergroup relations: Taking stock and looking forward. European review of
social psychology, 17(1), 271-320.
Pratto, F., & Stewart, A. L. (2012). Group dominance and the half‐ blindness of
privilege. Journal of Social Issues, 68(1), 28-45.
Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social dominance
orientation: A personality variable predicting social and political
attitudes. Journal of Personality and Social Psychology, 67(4), 741-763.
Ray, J. J. (1983). A scale to measure conservatism of American public opinion. Journal
of Social Psychology, 119(2), 293-294.
Riccio, M., Cole, S., & Balcetis, E. (2013). Seeing the expected, the desired, and the
feared: influences on perceptual interpretation and directed attention. Social and
Personality Psychology Compass, 7(6), 401-414.
Ross, L., Greene, D., & House, P. (1977). The “false consensus effect”: An egocentric
bias in social perception and attribution processes. Journal of Experimental Social
Psychology, 13(3), 279-301.
Sanitioso, R., Kunda, Z., & Fong, G. T. (1990). Motivated recruitment of
autobiographical memories. Journal of Personality and Social Psychology, 59,
229-241.
Schmitt, M. T., Branscombe, N. R., & Kappen, D. M. (2003). Attitudes toward group-
based inequality: Social dominance or social identity? British Journal of Social
Psychology, 42(2), 161-186.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
81
Sheehy-Skeffington, J. (2015). When hierarchy begins to hurt: Psychological
consequences of low subjective socioeconomic status at increasing levels of
inequality. Manuscript in preparation.
Sibley, C. G., & Liu, J. H. (2010). Social dominance orientation: Testing a global
individual difference perspective. Political Psychology, 31(2), 175-207.
Sibley, C. G., Wilson, M. S., & Duckitt, J. (2007). Antecedents of men’s hostile and
benevolent sexism: The dual roles of social dominance orientation and right-wing
authoritarianism. Personality and Social Psychology Bulletin, 33(2), 160-172.
Sidanius, J., Cotterill, S., Sheehy-Skeffington, J., Kteily, N.S., & Carvacho, H. (In Press).
Social Dominance Theory: Explorations in the Psychology of Oppression. In
Sibley, C. G. & Barlow, F. K. (Eds.), The Cambridge Handbook of the
Psychology of Prejudice.
Sidanius, J., & Pratto, F. (1999). Social dominance: An intergroup theory of social
hierarchy and oppression. Cambridge, UK: Cambridge University Press.
Sidanius, J., Pratto, F., Sinclair, S., & Van Laar, C. (1996). Mother Teresa meets Genghis
Khan: The dialectics of hierarchy-enhancing and hierarchy-attenuating career
choices. Social Justice Research, 9(2), 145-170.
Story, A. L. (1998). Self-esteem and memory for favorable and unfavorable personality
feedback. Personality and Social Psychology Bulletin, 24, 51-64.
Taber, C. S., & Lodge, M. (2006). Motivated skepticism in the evaluation of political
beliefs. American Journal of Political Science, 50(3), 755-769.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
82
Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W.
Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp.
33-48). Pacific Grove, CA: Brooks/Cole
Tetlock, P. E., & Mitchell, G. (1993). Liberal and conservative approaches to justice:
Conflicting psychological portraits. In Mellers, B. A. & Baron, J. (Eds.),
Psychological Perspectives on Justice. Theory and Applications (pp. 234-255).
New York, NY: Cambridge University Press.
Thomas, D. R., & Zumbo, B. D. (2012). Difference Scores From the Point of View of
Reliability and Repeated-Measures ANOVA In Defense of Difference Scores for
Data Analysis. Educational and Psychological Measurement, 72(1), 37-43.
Thomsen, L., Green, E. G., & Sidanius, J. (2008). We will hunt them down: How social
dominance orientation and right-wing authoritarianism fuel ethnic persecution of
immigrants in fundamentally different ways. Journal of Experimental Social
Psychology, 44(6), 1455-1464.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S.
(1987). Rediscovering the social group: A self-categorization theory. Basil
Blackwell.
Von Hippel, W., & Trivers, R. (2011). The evolution and psychology of self-
deception. Behavioral and Brain Sciences, 34(01), 1-16.
Waytz, A., Young, L. L., & Ginges, J. (2014). Motive attribution asymmetry for love vs.
hate drives intractable conflict. Proceedings of the National Academy of
Sciences, 111(44), 15687-15692.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
83
Williams, W. M., & Ceci, S. J. (2015). National hiring experiments reveal 2: 1 faculty
preference for women on STEM tenure track. Proceedings of the National
Academy of Sciences, 112(17), 5360-5365.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
84
Table 1a. Regression coefficients predicting support for overall power differences and social policies in Study 1a.
Note. With the exception of the indirect effects, all coefficients reflect standardized beta coefficients. *** p < .001 ** p < .01 * p < .05 Sex was coded 0= female, 1=male; Race was coded 0=White; 1= Black.
Overall Power
Differences
Social Welfare Support
Anti-Discriminatory Policy Support
Legacy Admissions
New Hierarchy
Affirmative Action Support
Overall
Egalitarian Policy
Support
SDO (logged) -.36*** -.45*** -.50*** .20*** .20*** -.03 -.47***
Social Conservatism -.10** -.19*** -.11** .03 -.01 -.11* -.11***
Social Class -.01 -.04 -.07* .09* .04 -.00 -.09**
Race -.15*** -.07* -.01 .16*** .05 .27*** -.10***
Sex -.08* -.03 -.03 -.00 .10* -.08* -.05
Overall Power Differences
-- .20*** .24*** -.17*** -.20*** -.03 .28***
Indirect effect
(SDO Power Differences
Policy)
-- -.19 [-.29, -.12] -.23 [-.31, -.15] .18 [.09, .29]
.19 [.09, .31]
.02 [-.05, .11] -.20 [-.27, -.14]
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
85
Figure 1. Example of an image of a fictitious society used in Study 3a.
The image shown is the median image in terms of degree of hierarchy/inequality.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
86
Figure 2. Example of an image of a fictitious organization used in Study 3b.
The image shown is the median image in terms of degree of hierarchy/inequality.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
87
Figure 3. An example of one of the four image sets in Study 5. The middle image here is the focal image (i.e., the actual image participants saw earlier in the study), with the top two images reflecting less hierarchical and the bottom two images more hierarchical versions. For each array, participants were presented with these images in randomized order.
RUNNING HEAD: HIERARCHY IN THE EYE OF THE BEHOLDER
88
Figure 4. Stylized example of the theoretical pattern of motivated biases in perception of hierarchy. The upper panel indicates predicted ratings of perceived hierarchy as a function of individuals’ SDO and the degree of hierarchy of the organization being rated. The lower panel plots this same information from the perspective of low and high SDO individuals’ deviations (in absolute value) from the theoretical ratings of those moderate in SDO.
1
2
3
4
5
6
7
Lesshierarchicalorganization
Averageorganization
Morehierarchicalorganization
Perc
epti
on o
f Hie
rarc
hy
Low SDO
Average SDO
High SDO
00.20.40.60.8
11.21.41.6
Less hierarchicalorganization
Averageorganization
Morehierarchicalorganization
Diff
eren
ce fr
om A
vera
ge S
DO
Low SDO
High SDO